🎁 Get the FREE AI Skills Starter Guide β€” Subscribe β†’
BytesAgainBytesAgain
BytesAgain β€Ί Skills β€Ί SOTA Agent
πŸ¦€ ClawHub

SOTA Agent

by @zack-dev-cm

SOTA Agent is a public ClawHub SOTA-campaign skill for CV and DS work. Use it when the user says "sota agent", "state of the art benchmark scouting", or want...

Versionv1.4.4
Downloads939
Stars⭐ 2
#legal#hr
TERMINAL
clawhub install sota-agent

πŸ“– About This Skill

name: sota-agent description: SOTA Agent is a public ClawHub SOTA-campaign skill for CV and DS work. Use it when the user says "sota agent", "state of the art benchmark scouting", or wants benchmark planning, paper triage, ablation design, and claim review for CV or data-science campaigns. version: 1.4.4 homepage: https://zack-dev-cm.github.io/ user-invocable: true metadata: {"openclaw":{"homepage":"https://zack-dev-cm.github.io/","skillKey":"sota-agent","requires":{"anyBins":["python3","python"]}}}

SOTA Agent

Search intent: sota agent, state of the art benchmark scouting, cv benchmark campaign, gpu vm research workflow

Goal

Turn a vague "beat the benchmark" request into a disciplined campaign:

  • fixed target metric and split
  • explicit literature and leaderboard snapshot
  • bounded reproduction plan
  • explicit handoff to the separate execution lane when runs need external tools
  • evidence requirements that can be reviewed without relying on live session state
  • ablations that answer one question at a time
  • promotion only when the claim survives review

    • This skill is the frontier-planning and candidate-selection layer. For execution artifacts or promotion evidence, pair it with data-science-cv-repro-lab; this skill stays focused on planning and claim review.

    Use This Skill When

  • the user wants a CV or DS system pushed toward state-of-the-art results
  • the task involves reproducing or surpassing recent papers
  • the workflow needs paper triage, leaderboard tracking, or claim review
  • the workflow needs a clean handoff to an execution skill after the benchmark contract is frozen
  • the user needs experiment management across local runs, notebooks, and long-running jobs
  • the question is whether execution evidence supports a SOTA candidate
  • the question is whether a candidate is a real SOTA step or only noise, leakage, or benchmark overfitting

    • If the campaign includes serious execution or release review, use this skill to choose and rank candidates, then use data-science-cv-repro-lab as the execution lane.

    Quick Start

    1. Freeze the claim target before touching recipes. - Name the task, dataset, metric, split, and target score. - Name the current trusted baseline. - Name the claim threshold for "match", "beat", or "not enough".

    2. Initialize the campaign records immediately. - Use python3 {baseDir}/scripts/init_sota_campaign.py --root

    --campaign-id --title </code>. - Use <code style="background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em">python3 {baseDir}/scripts/init_sota_leaderboard_snapshot.py --out <json> --task <task> --dataset <dataset> --metric <metric> --split <split></code>. - Use <code style="background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em">python3 {baseDir}/scripts/init_sota_paper_triage.py --out <json> --campaign-id <id> --task <task></code>. - Use <code style="background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em">python3 {baseDir}/scripts/init_sota_program.py --out <json> --campaign-id <id> --task <task> --dataset <dataset> --metric <metric> --split <split></code> when you need one machine-readable benchmark, rerun, delegation, and auth plan. - Use <code style="background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em">python3 {baseDir}/scripts/init_sota_candidate_card.py --out <json> --candidate-id <id> --campaign-id <id> --objective <goal></code>. - If execution review depends on synced QA runs, runtime sweeps, or benchmark panels, store the paired <code style="background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em">data-science-cv-repro-lab</code> review dashboard path in the program and candidate records before the claim review starts. - If external execution evidence exists, record the reviewed artifact manifest path in the program and candidate records instead of acting through a live session. - If the review surface needs manual or visual QA, use <code style="background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em">python3 {baseDir}/scripts/init_sota_validation_scorecard.py --out <json> --scorecard-id <id> --surface <surface></code>. - If an external export bundle matters, use <code style="background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em">python3 {baseDir}/scripts/init_sota_artifact_manifest.py --out <json> --bundle-root <dir></code>. - If a long execution run is involved, record only sanitized summaries and artifact references in the SOTA campaign files.</p><p style="margin:8px 0">3. Separate the campaign roles even if one agent performs all of them. - Scout: papers, leaderboards, repos, and benchmark rules. - Reproducer: baseline and top-paper reproduction. - Ablator: controlled change sets and compute allocation. - Reviewer: contamination, metric drift, and claim integrity. - Promoter: final claim or hold decision. - Keep the benchmark definition and final claim wording fixed. - Use bounded scouting and review lanes for literature triage, repo inspection, per-paper extraction, and hard-case review. - For repeated audits, batch over a manifest or CSV instead of free-form context accumulation.</p><p style="margin:8px 0">4. Pick the execution lane explicitly. - Execution handoff lane: use <code style="background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em">data-science-cv-repro-lab</code> for external runs and artifact capture. - Local lane: cheap falsification, tiny reruns, and artifact review.</p><p style="margin:8px 0">5. Keep file writes inside one campaign workspace. - Create one dedicated campaign root and keep every <code style="background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em">--out</code>, <code style="background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em">--bundle-root</code>, and <code style="background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em">--output-root</code> path under it. - Do not point the bundled scripts at unrelated home-directory or system paths. - Treat <code style="background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em">scripts/sota_public_safety.py</code> as the canonical public-redaction layer for URLs, refs, and paths.</p><p style="margin:8px 0">6. Work the SOTA ladder in order. - Freeze the benchmark definition and auth rule before using more compute. - Reproduce the trusted baseline first. - Reproduce one relevant reference result or a close public checkpoint. - Build a hypothesis backlog from literature gaps, not vibes. - Run narrow ablations before broad recipe churn. - Stress the best candidate on the fixed review surfaces.</p><p style="margin:8px 0">7. Claim only on full-surface wins. - Fixed benchmark score - Reproduced baseline delta - Compute or cost context - Browser or GUI evidence if that lane mattered - Failure-case review - Exact evidence bundle - Render the final review with <code style="background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em">python3 {baseDir}/scripts/render_sota_claim_summary.py --candidate-card <json> --out <md></code>.</p><p style="margin:8px 0"><h3 style="color:#e5e7eb;margin:18px 0 8px;font-size:1.05em">Operating Rules</h3></p><p style="margin:8px 0"><h4 style="color:#d1d5db;margin:14px 0 6px;font-size:.95em">Campaign rules</h4></p><p style="margin:8px 0"><li style="color:#94a3b8;margin:3px 0">One campaign has one target benchmark contract.</li> <li style="color:#94a3b8;margin:3px 0">Do not let the target metric or split drift midstream.</li> <li style="color:#94a3b8;margin:3px 0">Keep a short hypothesis backlog and kill low-information ideas quickly.</li> <li style="color:#94a3b8;margin:3px 0">Record why each experiment exists before running it.</li></p><p style="margin:8px 0"><h4 style="color:#d1d5db;margin:14px 0 6px;font-size:.95em">Codex multi-agent rules</h4></p><p style="margin:8px 0"><li style="color:#94a3b8;margin:3px 0">Main thread owns the benchmark contract, stop conditions, and final claim decision.</li> <li style="color:#94a3b8;margin:3px 0">Subagents should do bounded work only: scout, reproduce, ablate, or review.</li> <li style="color:#94a3b8;margin:3px 0">Do not let one exploratory thread silently rewrite the campaign contract.</li> <li style="color:#94a3b8;margin:3px 0">For repeated claim checks or literature extraction, prefer manifest-driven fanout over conversational drift.</li></p><p style="margin:8px 0"><h4 style="color:#d1d5db;margin:14px 0 6px;font-size:.95em">Literature rules</h4></p><p style="margin:8px 0"><li style="color:#94a3b8;margin:3px 0">Read only the papers or repos that change the candidate plan.</li> <li style="color:#94a3b8;margin:3px 0">Extract the minimum useful fields: task, metric, split, data, compute, architecture, augmentations, training tricks, and caveats.</li> <li style="color:#94a3b8;margin:3px 0">Prefer a reproduced strong baseline over copying five tricks from five papers without control.</li> <li style="color:#94a3b8;margin:3px 0">Do not treat leaderboard rows as ground truth without checking task definition and split rules.</li></p><p style="margin:8px 0"><h4 style="color:#d1d5db;margin:14px 0 6px;font-size:.95em">Ablation rules</h4></p><p style="margin:8px 0"><li style="color:#94a3b8;margin:3px 0">Change one meaningful variable at a time when the goal is causal understanding.</li> <li style="color:#94a3b8;margin:3px 0">If several knobs move together, label the run as a package change, not an ablation.</li> <li style="color:#94a3b8;margin:3px 0">Keep one canonical baseline recipe alive for comparison.</li> <li style="color:#94a3b8;margin:3px 0">Require the first winning candidate to survive at least one rerun or adjacent-seed check before escalating the claim.</li></p><p style="margin:8px 0"><h4 style="color:#d1d5db;margin:14px 0 6px;font-size:.95em">Compute rules</h4></p><p style="margin:8px 0"><li style="color:#94a3b8;margin:3px 0">Spend cheap compute on reproduction and short falsification first.</li> <li style="color:#94a3b8;margin:3px 0">Do not push a long run unless the hypothesis would matter if it wins.</li> <li style="color:#94a3b8;margin:3px 0">Record training cost, wall time, and hardware for every serious candidate.</li> <li style="color:#94a3b8;margin:3px 0">Cut branches that cannot plausibly clear the target with the remaining budget.</li></p><p style="margin:8px 0"><h4 style="color:#d1d5db;margin:14px 0 6px;font-size:.95em">Runtime and auth rules</h4></p><p style="margin:8px 0"><li style="color:#94a3b8;margin:3px 0">This public skill does not require API keys, account tokens, live sessions, or account-bound credentials.</li> <li style="color:#94a3b8;margin:3px 0">Prefer local files, public URLs, and user-supplied artifacts over account-bound execution paths.</li> <li style="color:#94a3b8;margin:3px 0">Do not require or recommend <code style="background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em">OPENAI_API_KEY</code>, other vendor API keys, or paid inference APIs as the default campaign runtime path.</li> <li style="color:#94a3b8;margin:3px 0">If a third-party framework only works through paid API keys, treat it as reference material unless it can run through local tools or public artifacts.</li></p><p style="margin:8px 0"><h4 style="color:#d1d5db;margin:14px 0 6px;font-size:.95em">External execution rules</h4></p><p style="margin:8px 0"><li style="color:#94a3b8;margin:3px 0">Execution rules live in <code style="background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em">data-science-cv-repro-lab</code>.</li> <li style="color:#94a3b8;margin:3px 0">In this skill, record only the benchmark contract, candidate rationale, review status, and sanitized artifact references.</li></p><p style="margin:8px 0"><h4 style="color:#d1d5db;margin:14px 0 6px;font-size:.95em">Claim safety rules</h4></p><p style="margin:8px 0"><li style="color:#94a3b8;margin:3px 0">No SOTA claim without a fixed metric, split, and baseline.</li> <li style="color:#94a3b8;margin:3px 0">No SOTA claim on a contaminated benchmark or hidden train-on-test path.</li> <li style="color:#94a3b8;margin:3px 0">If the execution story depends on a dashboard or synced review surface, keep the dashboard path, source audit, and leakage audit in the claim packet.</li> <li style="color:#94a3b8;margin:3px 0">If a candidate wins only on one slice while regressing important surfaces, hold it.</li> <li style="color:#94a3b8;margin:3px 0">Report uncertainty honestly: "best internal result so far" is not the same as "new SOTA".</li> <li style="color:#94a3b8;margin:3px 0">Small deltas need rerun or adjacent-seed support before they become claim language.</li></p><p style="margin:8px 0"><h3 style="color:#e5e7eb;margin:18px 0 8px;font-size:1.05em">References</h3></p><p style="margin:8px 0">Read only the reference that matches the task:</p><p style="margin:8px 0"><li style="color:#94a3b8;margin:3px 0"><code style="background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em">references/sota-campaign-playbook.md</code></li> - Full campaign structure, role separation, and stop conditions. <li style="color:#94a3b8;margin:3px 0"><code style="background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em">references/sota-program-rules.md</code></li> - Rules for queues, stage discipline, ablations, and promotion gating. <li style="color:#94a3b8;margin:3px 0"><code style="background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em">references/campaign-harness-stack.md</code></li> - What to reuse from Codex subagents, harness engineering, OpenEvolve, Symphony, Paperclip, and OptiLLM under a local-first campaign rule. <li style="color:#94a3b8;margin:3px 0"><code style="background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em">references/benchmark-discipline.md</code></li> - How to avoid contamination, metric drift, and invalid comparisons. <li style="color:#94a3b8;margin:3px 0"><code style="background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em">references/paper-triage.md</code></li> - How to filter papers and extract only decision-relevant details. <li style="color:#94a3b8;margin:3px 0"><code style="background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em">references/public-research-lane.md</code></li> - How to review public literature and leaderboard pages without private sessions. <li style="color:#94a3b8;margin:3px 0"><code style="background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em">references/external-evidence-handoff.md</code></li> - How to record sanitized evidence from external notebook or UI runs without controlling a live session. <li style="color:#94a3b8;margin:3px 0"><code style="background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em">references/execution-evidence-summary.md</code></li> - How to summarize execution evidence that belongs in the paired execution skill. <li style="color:#94a3b8;margin:3px 0"><code style="background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em">references/claim-safety.md</code></li> - Review rules for whether a candidate deserves a SOTA claim at all. <li style="color:#94a3b8;margin:3px 0"><code style="background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em">references/public-safety.md</code></li> - Publication review rules for secrets, private refs, and raw notebook paths.</p><p style="margin:8px 0"><h3 style="color:#e5e7eb;margin:18px 0 8px;font-size:1.05em">Bundled Scripts</h3></p><p style="margin:8px 0"><li style="color:#94a3b8;margin:3px 0"><code style="background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em">scripts/sota_public_safety.py</code></li> - Pure local helpers for path, URL, ref, env, and command redaction. No network I/O or subprocess execution. <li style="color:#94a3b8;margin:3px 0"><code style="background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em">scripts/init_sota_campaign.py</code></li> - Create a reusable campaign folder with benchmark, program, agent, research, leaderboard, plan, ablation, evidence, and claim files. <li style="color:#94a3b8;margin:3px 0"><code style="background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em">scripts/init_sota_program.py</code></li> - Create a machine-readable program record with the fixed benchmark, baselines, rerun policy, bounded subagent roles, and local-first runtime rules. <li style="color:#94a3b8;margin:3px 0"><code style="background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em">scripts/init_sota_leaderboard_snapshot.py</code></li> - Create a machine-readable snapshot of the target benchmark contract and current reference scores. <li style="color:#94a3b8;margin:3px 0"><code style="background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em">scripts/init_sota_paper_triage.py</code></li> - Create a machine-readable literature queue for paper screening and extraction. <li style="color:#94a3b8;margin:3px 0"><code style="background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em">scripts/init_sota_browser_run_card.py</code></li> - Create a sanitized external-evidence record for notebook or UI run artifacts. <li style="color:#94a3b8;margin:3px 0"><code style="background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em">scripts/init_sota_validation_scorecard.py</code></li> - Create a machine-readable GUI or notebook validation scorecard when visible state matters to the campaign. <li style="color:#94a3b8;margin:3px 0"><code style="background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em">scripts/init_sota_artifact_manifest.py</code></li> - Create a machine-readable export-bundle manifest for external artifacts with redacted public path metadata. <li style="color:#94a3b8;margin:3px 0"><code style="background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em">scripts/init_sota_candidate_card.py</code></li> - Create a machine-readable card for a serious candidate, its execution lane, auth mode, and claim state. <li style="color:#94a3b8;margin:3px 0"><code style="background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em">scripts/init_sota_candidate.py</code></li> - Create a machine-readable candidate record with change set, risks, and redacted public artifact refs. <li style="color:#94a3b8;margin:3px 0"><code style="background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em">scripts/init_sota_ablation_queue.py</code></li> - Create a focused ablation queue for one candidate family. <li style="color:#94a3b8;margin:3px 0"><code style="background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em">scripts/init_sota_vm_bootstrap_manifest.py</code></li> - Create a redacted long-run summary manifest for already-approved execution artifacts. <li style="color:#94a3b8;margin:3px 0"><code style="background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em">scripts/update_sota_scoreboard.py</code></li> - Refresh a ranked scoreboard for a fixed metric and goal direction. <li style="color:#94a3b8;margin:3px 0"><code style="background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em">scripts/init_sota_review_packet.py</code></li> - Join the core artifacts for a promotion, hold, or cut decision. <li style="color:#94a3b8;margin:3px 0"><code style="background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em">scripts/render_sota_claim_summary.py</code></li> - Render a concise markdown review from the machine-readable candidate card. <li style="color:#94a3b8;margin:3px 0"><code style="background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em">scripts/render_sota_program_summary.py</code></li> - Render a concise markdown summary from the program, candidate, scoreboard, and review packet. </p></div></section><section class="skill-card" style="margin-bottom:20px"><h2 style="color:#f8fafc;font-size:1.2em;font-weight:800;margin:0 0 16px;display:flex;align-items:center;gap:8px">πŸ’‘ Examples</h2><div style="font-size:.92em;color:#94a3b8;line-height:1.75"><p style="margin:8px 0">1. Freeze the claim target before touching recipes. - Name the task, dataset, metric, split, and target score. - Name the current trusted baseline. - Name the claim threshold for "match", "beat", or "not enough".</p><p style="margin:8px 0">2. Initialize the campaign records immediately. - Use <code style="background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em">python3 {baseDir}/scripts/init_sota_campaign.py --root <dir> --campaign-id <id> --title <title></code>. - Use <code style="background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em">python3 {baseDir}/scripts/init_sota_leaderboard_snapshot.py --out <json> --task <task> --dataset <dataset> --metric <metric> --split <split></code>. - Use <code style="background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em">python3 {baseDir}/scripts/init_sota_paper_triage.py --out <json> --campaign-id <id> --task <task></code>. - Use <code style="background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em">python3 {baseDir}/scripts/init_sota_program.py --out <json> --campaign-id <id> --task <task> --dataset <dataset> --metric <metric> --split <split></code> when you need one machine-readable benchmark, rerun, delegation, and auth plan. - Use <code style="background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em">python3 {baseDir}/scripts/init_sota_candidate_card.py --out <json> --candidate-id <id> --campaign-id <id> --objective <goal></code>. - If execution review depends on synced QA runs, runtime sweeps, or benchmark panels, store the paired <code style="background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em">data-science-cv-repro-lab</code> review dashboard path in the program and candidate records before the claim review starts. - If external execution evidence exists, record the reviewed artifact manifest path in the program and candidate records instead of acting through a live session. - If the review surface needs manual or visual QA, use <code style="background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em">python3 {baseDir}/scripts/init_sota_validation_scorecard.py --out <json> --scorecard-id <id> --surface <surface></code>. - If an external export bundle matters, use <code style="background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em">python3 {baseDir}/scripts/init_sota_artifact_manifest.py --out <json> --bundle-root <dir></code>. - If a long execution run is involved, record only sanitized summaries and artifact references in the SOTA campaign files.</p><p style="margin:8px 0">3. Separate the campaign roles even if one agent performs all of them. - Scout: papers, leaderboards, repos, and benchmark rules. - Reproducer: baseline and top-paper reproduction. - Ablator: controlled change sets and compute allocation. - Reviewer: contamination, metric drift, and claim integrity. - Promoter: final claim or hold decision. - Keep the benchmark definition and final claim wording fixed. - Use bounded scouting and review lanes for literature triage, repo inspection, per-paper extraction, and hard-case review. - For repeated audits, batch over a manifest or CSV instead of free-form context accumulation.</p><p style="margin:8px 0">4. Pick the execution lane explicitly. - Execution handoff lane: use <code style="background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em">data-science-cv-repro-lab</code> for external runs and artifact capture. - Local lane: cheap falsification, tiny reruns, and artifact review.</p><p style="margin:8px 0">5. Keep file writes inside one campaign workspace. - Create one dedicated campaign root and keep every <code style="background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em">--out</code>, <code style="background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em">--bundle-root</code>, and <code style="background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em">--output-root</code> path under it. - Do not point the bundled scripts at unrelated home-directory or system paths. - Treat <code style="background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em">scripts/sota_public_safety.py</code> as the canonical public-redaction layer for URLs, refs, and paths.</p><p style="margin:8px 0">6. Work the SOTA ladder in order. - Freeze the benchmark definition and auth rule before using more compute. - Reproduce the trusted baseline first. - Reproduce one relevant reference result or a close public checkpoint. - Build a hypothesis backlog from literature gaps, not vibes. - Run narrow ablations before broad recipe churn. - Stress the best candidate on the fixed review surfaces.</p><p style="margin:8px 0">7. Claim only on full-surface wins. - Fixed benchmark score - Reproduced baseline delta - Compute or cost context - Browser or GUI evidence if that lane mattered - Failure-case review - Exact evidence bundle - Render the final review with <code style="background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em">python3 {baseDir}/scripts/render_sota_claim_summary.py --candidate-card <json> --out <md></code>.</p></div></section></div><div class="two-col-side"></div></div></div><script> document.querySelectorAll('.copy-btn, .script-copy-btn').forEach(btn => { btn.addEventListener('click', () => { const cmd = btn.getAttribute('data-cmd'); if (!cmd) return; navigator.clipboard.writeText(cmd).then(() => { const orig = btn.textContent; btn.textContent = 'Copied!'; setTimeout(() => btn.textContent = orig, 1500); }).catch(() => {}); }); }); </script><!--$--><!--/$--></main><footer style="background:var(--bg-primary);border-top:1px solid var(--border-secondary);margin-top:60px"><div style="border-top:1px solid var(--border-light);max-width:1200px;margin:0 auto;padding:24px 20px"><div style="display:flex;justify-content:space-between;flex-wrap:wrap;gap:24px;margin-bottom:24px"><div><div style="font-weight:700;color:var(--text-muted);margin-bottom:8px">BytesAgain</div><div style="color:var(--text-muted3);font-size:.82em;max-width:200px">Discover the best AI agent skills for your workflow.</div></div><div><div style="color:var(--text-muted);font-size:.75em;text-transform:uppercase;letter-spacing:1px;margin-bottom:10px">Explore</div><div style="margin-bottom:6px"><a href="/skills" style="color:var(--text-muted2);text-decoration:none;font-size:.85em">Skills</a></div><div style="margin-bottom:6px"><a href="/articles" style="color:var(--text-muted2);text-decoration:none;font-size:.85em">Articles</a></div><div style="margin-bottom:6px"><a href="/use-case" style="color:var(--text-muted2);text-decoration:none;font-size:.85em">Cases</a></div></div><div><div style="color:var(--text-muted);font-size:.75em;text-transform:uppercase;letter-spacing:1px;margin-bottom:10px">Company</div><div style="margin-bottom:6px"><a href="/about" style="color:var(--text-muted2);text-decoration:none;font-size:.85em">About</a></div><div style="margin-bottom:6px"><a href="/contact" style="color:var(--text-muted2);text-decoration:none;font-size:.85em">Contact</a></div><div style="margin-bottom:6px"><a href="/privacy-policy" style="color:var(--text-muted2);text-decoration:none;font-size:.85em">Privacy Policy</a></div><div style="margin-bottom:6px"><a href="/terms" style="color:var(--text-muted2);text-decoration:none;font-size:.85em">Terms</a></div><div style="margin-bottom:6px"><a href="/feedback" style="color:var(--text-muted2);text-decoration:none;font-size:.85em">Feedback</a></div></div></div><div style="border-top:1px solid var(--border-light);padding-top:16px"><div style="color:var(--text-muted4);font-size:.8em;margin-bottom:8px">Β© <!-- -->2026<!-- --> BytesAgain. All rights reserved.</div><div style="color:var(--text-muted5);font-size:.75em;line-height:1.6;max-width:720px">BytesAgain is an independent skill directory. We index and link to third-party content (ClawHub, GitHub, LobeHub, Dify, etc.) for informational purposes only. All trademarks, skill names, and content are the property of their respective owners. BytesAgain does not claim ownership of any indexed content.</div></div></div></footer><button style="position:fixed;bottom:28px;right:28px;z-index:1000;width:48px;height:48px;border-radius:50%;border:none;cursor:pointer;background:linear-gradient(135deg,#667eea,#00d4ff);color:#fff;font-size:1.3em;box-shadow:0 4px 20px #667eea66;display:flex;align-items:center;justify-content:center;transition:transform .2s">πŸ’¬</button><script src="/_next/static/chunks/0ze4gu236oq96.js?dpl=dpl_9zPFH6pojPkqyspdDNz28GzCc6KQ" id="_R_" async=""></script><script>(self.__next_f=self.__next_f||[]).push([0])</script><script>self.__next_f.push([1,"1:\"$Sreact.fragment\"\n2:I[62894,[\"/_next/static/chunks/0ph_0rx9ah5dc.js?dpl=dpl_9zPFH6pojPkqyspdDNz28GzCc6KQ\",\"/_next/static/chunks/0i_x3w546rsb3.js?dpl=dpl_9zPFH6pojPkqyspdDNz28GzCc6KQ\",\"/_next/static/chunks/06ig5gym-0n-u.js?dpl=dpl_9zPFH6pojPkqyspdDNz28GzCc6KQ\"],\"LangProvider\"]\n3:I[89220,[\"/_next/static/chunks/0ph_0rx9ah5dc.js?dpl=dpl_9zPFH6pojPkqyspdDNz28GzCc6KQ\",\"/_next/static/chunks/0i_x3w546rsb3.js?dpl=dpl_9zPFH6pojPkqyspdDNz28GzCc6KQ\",\"/_next/static/chunks/06ig5gym-0n-u.js?dpl=dpl_9zPFH6pojPkqyspdDNz28GzCc6KQ\"],\"ThemeProvider\"]\n4:I[16988,[\"/_next/static/chunks/0ph_0rx9ah5dc.js?dpl=dpl_9zPFH6pojPkqyspdDNz28GzCc6KQ\",\"/_next/static/chunks/0i_x3w546rsb3.js?dpl=dpl_9zPFH6pojPkqyspdDNz28GzCc6KQ\",\"/_next/static/chunks/06ig5gym-0n-u.js?dpl=dpl_9zPFH6pojPkqyspdDNz28GzCc6KQ\"],\"default\"]\ne:I[68027,[\"/_next/static/chunks/0ph_0rx9ah5dc.js?dpl=dpl_9zPFH6pojPkqyspdDNz28GzCc6KQ\",\"/_next/static/chunks/0i_x3w546rsb3.js?dpl=dpl_9zPFH6pojPkqyspdDNz28GzCc6KQ\",\"/_next/static/chunks/06ig5gym-0n-u.js?dpl=dpl_9zPFH6pojPkqyspdDNz28GzCc6KQ\"],\"default\",1]\n:HL[\"/_next/static/chunks/051nc0vy_6.rl.css?dpl=dpl_9zPFH6pojPkqyspdDNz28GzCc6KQ\",\"style\"]\n:HL[\"/_next/static/media/caa3a2e1cccd8315-s.p.09~u27dqhyhd6.woff2?dpl=dpl_9zPFH6pojPkqyspdDNz28GzCc6KQ\",\"font\",{\"crossOrigin\":\"\",\"type\":\"font/woff2\"}]\n5:Td5e,"])</script><script>self.__next_f.push([1,"[{\"@context\":\"https://schema.org\",\"@type\":\"WebSite\",\"name\":\"BytesAgain\",\"url\":\"https://bytesagain.com\",\"description\":\"Search 60,000+ verified AI agent skills via MCP API or REST. Supports 7 languages. Free, no auth required.\",\"inLanguage\":[\"en\",\"zh\",\"es\",\"fr\",\"de\",\"ja\",\"ko\"],\"potentialAction\":{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https://bytesagain.com/skills?q={search_term_string}\"},\"query-input\":\"required name=search_term_string\"}},{\"@context\":\"https://schema.org\",\"@type\":\"Organization\",\"name\":\"BytesAgain\",\"url\":\"https://bytesagain.com\",\"logo\":{\"@type\":\"ImageObject\",\"url\":\"https://bytesagain.com/og-image.png\"},\"description\":\"AI agent skill directory. Search 60,000+ skills, 1,000+ use cases, and community requests.\",\"foundingDate\":\"2026\",\"foundingLocation\":{\"@type\":\"Place\",\"name\":\"Global\"},\"sameAs\":[\"https://x.com/bytesagain\",\"https://github.com/bytesagain/ai-skills\",\"https://clawhub.ai/profile/bytesagain\"],\"contactPoint\":{\"@type\":\"ContactPoint\",\"email\":\"hello@bytesagain.com\",\"contactType\":\"customer support\"},\"numberOfEmployees\":{\"@type\":\"QuantitativeValue\",\"value\":1}},{\"@context\":\"https://schema.org\",\"@type\":\"WebApplication\",\"name\":\"BytesAgain AI Skills Search\",\"url\":\"https://bytesagain.com\",\"applicationCategory\":\"DeveloperApplication\",\"operatingSystem\":\"Web\",\"description\":\"Search engine and MCP API for 60,000+ AI agent skills. Semantic search, role recommendations, and use case packs.\",\"offers\":{\"@type\":\"Offer\",\"price\":\"0\",\"priceCurrency\":\"USD\"},\"featureList\":[\"Search 60,000+ AI agent skills\",\"Role-based recommendations for developers, creators, and traders\",\"1,000+ curated use case packs\",\"Free MCP API and REST API\",\"Multi-language search (EN, ZH, ES, FR, DE, JA, KO)\"],\"potentialAction\":{\"@type\":\"SearchAction\",\"target\":\"https://bytesagain.com/skills?q={search_term_string}\",\"query-input\":\"required name=search_term_string\"},\"dateModified\":\"2026-08-02\"},{\"@context\":\"https://schema.org\",\"@type\":\"FAQPage\",\"mainEntity\":[{\"@type\":\"Question\",\"name\":\"What is BytesAgain?\",\"acceptedAnswer\":{\"@type\":\"Answer\",\"text\":\"BytesAgain is a curated directory of 60,000+ AI agent skills from ClawHub, GitHub, LobeHub, and Dify. Search skills by keyword in 7 languages, browse by role (developer, creator, trader, marketer) or by use case.\"}},{\"@type\":\"Question\",\"name\":\"How do I find AI skills on BytesAgain?\",\"acceptedAnswer\":{\"@type\":\"Answer\",\"text\":\"Use the search bar on BytesAgain.com to search by keyword in 7 languages. You can also browse by role (developer, creator, trader, marketer) or by use case. Each skill shows install instructions for Claude, Cursor, OpenClaw, Continue, and more.\"}},{\"@type\":\"Question\",\"name\":\"Is BytesAgain free?\",\"acceptedAnswer\":{\"@type\":\"Answer\",\"text\":\"Yes, BytesAgain is completely free. No registration required for searching skills. The MCP API is also free with rate limits.\"}},{\"@type\":\"Question\",\"name\":\"Does BytesAgain have an API for AI agents?\",\"acceptedAnswer\":{\"@type\":\"Answer\",\"text\":\"Yes! BytesAgain provides a free MCP SSE endpoint at /api/mcp/sse for AI agents, plus a REST API at /api/mcp?action=search\u0026q=\u003cquery\u003e. No authentication needed.\"}},{\"@type\":\"Question\",\"name\":\"Can I request a new AI skill on BytesAgain?\",\"acceptedAnswer\":{\"@type\":\"Answer\",\"text\":\"Yes! Visit the Requests page on BytesAgain.com to submit a skill request. Your request will be visible to the community and notified to the site admin.\"}}]}]"])</script><script>self.__next_f.push([1,"0:{\"P\":null,\"c\":[\"\",\"skill\",\"sota-agent\"],\"q\":\"\",\"i\":false,\"f\":[[[\"\",{\"children\":[\"skill\",{\"children\":[[\"slug\",\"sota-agent\",\"d\",null],{\"children\":[\"__PAGE__\",{}]}]}]},\"$undefined\",\"$undefined\",16],[[\"$\",\"$1\",\"c\",{\"children\":[[[\"$\",\"link\",\"0\",{\"rel\":\"stylesheet\",\"href\":\"/_next/static/chunks/051nc0vy_6.rl.css?dpl=dpl_9zPFH6pojPkqyspdDNz28GzCc6KQ\",\"precedence\":\"next\",\"crossOrigin\":\"$undefined\",\"nonce\":\"$undefined\"}],[\"$\",\"script\",\"script-0\",{\"src\":\"/_next/static/chunks/0ph_0rx9ah5dc.js?dpl=dpl_9zPFH6pojPkqyspdDNz28GzCc6KQ\",\"async\":true,\"nonce\":\"$undefined\"}],[\"$\",\"script\",\"script-1\",{\"src\":\"/_next/static/chunks/0i_x3w546rsb3.js?dpl=dpl_9zPFH6pojPkqyspdDNz28GzCc6KQ\",\"async\":true,\"nonce\":\"$undefined\"}],[\"$\",\"script\",\"script-2\",{\"src\":\"/_next/static/chunks/06ig5gym-0n-u.js?dpl=dpl_9zPFH6pojPkqyspdDNz28GzCc6KQ\",\"async\":true,\"nonce\":\"$undefined\"}]],[\"$\",\"html\",null,{\"lang\":\"en\",\"children\":[[\"$\",\"head\",null,{\"children\":[[\"$\",\"link\",null,{\"rel\":\"llms\",\"href\":\"/llms.txt\"}],[\"$\",\"link\",null,{\"rel\":\"llms-full\",\"href\":\"/llms-full.txt\"}],[\"$\",\"script\",null,{\"async\":true,\"src\":\"https://www.googletagmanager.com/gtag/js?id=G-3C1MM9FWYF\"}],[\"$\",\"script\",null,{\"dangerouslySetInnerHTML\":{\"__html\":\"\\n window.dataLayer = window.dataLayer || [];\\n function gtag(){dataLayer.push(arguments);}\\n gtag('js', new Date());\\n gtag('config', 'G-3C1MM9FWYF');\\n \"}}]]}],[\"$\",\"body\",null,{\"className\":\"geist_9e050971-module__05dp7a__className\",\"style\":{\"margin\":0},\"children\":[\"$\",\"$L2\",null,{\"children\":[\"$\",\"$L3\",null,{\"children\":[[\"$\",\"div\",null,{\"style\":{\"width\":\"100%\",\"background\":\"var(--bg-subscribe)\",\"borderBottom\":\"1px solid var(--border-primary)\",\"padding\":\"8px 20px\",\"textAlign\":\"center\",\"fontSize\":\".82em\",\"color\":\"#818cf8\"},\"children\":[\"🎁 \",[\"$\",\"strong\",null,{\"style\":{\"color\":\"var(--text-primary)\"},\"children\":\"Get the FREE AI Skills Starter Guide\"}],\" β€” \",[\"$\",\"a\",null,{\"href\":\"/register\",\"style\":{\"color\":\"#00d4ff\",\"textDecoration\":\"underline\"},\"children\":\"Subscribe β†’\"}]]}],[\"$\",\"$L4\",null,{}],[\"$\",\"script\",null,{\"type\":\"application/ld+json\",\"dangerouslySetInnerHTML\":{\"__html\":\"$5\"}}],\"$L6\",\"$L7\",\"$L8\"]}]}]}]]}]]}],{\"children\":[\"$L9\",{\"children\":[\"$La\",{\"children\":[\"$Lb\",{},null,false,null]},null,false,\"$@c\"]},null,false,\"$@c\"]},null,false,null],\"$Ld\",false]],\"m\":\"$undefined\",\"G\":[\"$e\",[\"$Lf\"]],\"S\":true,\"h\":null,\"s\":\"$undefined\",\"l\":\"$undefined\",\"p\":\"$undefined\",\"d\":\"$undefined\"}\n"])</script><script>self.__next_f.push([1,"10:I[39756,[\"/_next/static/chunks/0ph_0rx9ah5dc.js?dpl=dpl_9zPFH6pojPkqyspdDNz28GzCc6KQ\",\"/_next/static/chunks/0i_x3w546rsb3.js?dpl=dpl_9zPFH6pojPkqyspdDNz28GzCc6KQ\",\"/_next/static/chunks/06ig5gym-0n-u.js?dpl=dpl_9zPFH6pojPkqyspdDNz28GzCc6KQ\"],\"default\"]\n11:I[37457,[\"/_next/static/chunks/0ph_0rx9ah5dc.js?dpl=dpl_9zPFH6pojPkqyspdDNz28GzCc6KQ\",\"/_next/static/chunks/0i_x3w546rsb3.js?dpl=dpl_9zPFH6pojPkqyspdDNz28GzCc6KQ\",\"/_next/static/chunks/06ig5gym-0n-u.js?dpl=dpl_9zPFH6pojPkqyspdDNz28GzCc6KQ\"],\"default\"]\n12:I[22016,[\"/_next/static/chunks/0ph_0rx9ah5dc.js?dpl=dpl_9zPFH6pojPkqyspdDNz28GzCc6KQ\",\"/_next/static/chunks/0i_x3w546rsb3.js?dpl=dpl_9zPFH6pojPkqyspdDNz28GzCc6KQ\",\"/_next/static/chunks/06ig5gym-0n-u.js?dpl=dpl_9zPFH6pojPkqyspdDNz28GzCc6KQ\",\"/_next/static/chunks/0ka051yepewro.js?dpl=dpl_9zPFH6pojPkqyspdDNz28GzCc6KQ\"],\"\"]\n13:I[90940,[\"/_next/static/chunks/0ph_0rx9ah5dc.js?dpl=dpl_9zPFH6pojPkqyspdDNz28GzCc6KQ\",\"/_next/static/chunks/0i_x3w546rsb3.js?dpl=dpl_9zPFH6pojPkqyspdDNz28GzCc6KQ\",\"/_next/static/chunks/06ig5gym-0n-u.js?dpl=dpl_9zPFH6pojPkqyspdDNz28GzCc6KQ\"],\"default\"]\n14:I[16397,[\"/_next/static/chunks/0ph_0rx9ah5dc.js?dpl=dpl_9zPFH6pojPkqyspdDNz28GzCc6KQ\",\"/_next/static/chunks/0i_x3w546rsb3.js?dpl=dpl_9zPFH6pojPkqyspdDNz28GzCc6KQ\",\"/_next/static/chunks/06ig5gym-0n-u.js?dpl=dpl_9zPFH6pojPkqyspdDNz28GzCc6KQ\"],\"default\"]\n16:I[97367,[\"/_next/static/chunks/0ph_0rx9ah5dc.js?dpl=dpl_9zPFH6pojPkqyspdDNz28GzCc6KQ\",\"/_next/static/chunks/0i_x3w546rsb3.js?dpl=dpl_9zPFH6pojPkqyspdDNz28GzCc6KQ\",\"/_next/static/chunks/06ig5gym-0n-u.js?dpl=dpl_9zPFH6pojPkqyspdDNz28GzCc6KQ\"],\"OutletBoundary\"]\n17:\"$Sreact.suspense\"\n1a:I[97367,[\"/_next/static/chunks/0ph_0rx9ah5dc.js?dpl=dpl_9zPFH6pojPkqyspdDNz28GzCc6KQ\",\"/_next/static/chunks/0i_x3w546rsb3.js?dpl=dpl_9zPFH6pojPkqyspdDNz28GzCc6KQ\",\"/_next/static/chunks/06ig5gym-0n-u.js?dpl=dpl_9zPFH6pojPkqyspdDNz28GzCc6KQ\"],\"ViewportBoundary\"]\n1c:I[97367,[\"/_next/static/chunks/0ph_0rx9ah5dc.js?dpl=dpl_9zPFH6pojPkqyspdDNz28GzCc6KQ\",\"/_next/static/chunks/0i_x3w546rsb3.js?dpl=dpl_9zPFH6pojPkqyspdDNz28GzCc6KQ\",\"/_next/static/chunks/06ig5gym-0n-u.js?dpl=dpl_9zPFH6pojPkqyspdDNz28GzCc6KQ\"],\"MetadataBoundary\"]\n"])</script><script>self.__next_f.push([1,"6:[\"$\",\"main\",null,{\"children\":[\"$\",\"$L10\",null,{\"parallelRouterKey\":\"children\",\"error\":\"$undefined\",\"errorStyles\":\"$undefined\",\"errorScripts\":\"$undefined\",\"template\":[\"$\",\"$L11\",null,{}],\"templateStyles\":\"$undefined\",\"templateScripts\":\"$undefined\",\"notFound\":[[\"$\",\"main\",null,{\"style\":{\"minHeight\":\"100vh\",\"display\":\"flex\",\"alignItems\":\"center\",\"justifyContent\":\"center\",\"background\":\"#050611\",\"color\":\"#e5e7eb\"},\"children\":[[\"$\",\"style\",null,{\"children\":\"\\n .nf-box { text-align: center; padding: 60px 32px; }\\n .nf-code { font-size: 6rem; font-weight: 900; color: #22d3ee; line-height: 1; margin: 0; }\\n .nf-title { font-size: 1.8rem; font-weight: 800; margin: 12px 0 8px; }\\n .nf-desc { color: var(--text-muted2); font-size: 1rem; margin-bottom: 32px; max-width: 440px; }\\n .nf-link { display: inline-block; padding: 12px 28px; background: linear-gradient(135deg,#34d399,#22d3ee); color: #000; font-weight: 900; border-radius: 12px; text-decoration: none; }\\n \"}],[\"$\",\"div\",null,{\"className\":\"nf-box\",\"children\":[[\"$\",\"p\",null,{\"className\":\"nf-code\",\"children\":\"404\"}],[\"$\",\"h1\",null,{\"className\":\"nf-title\",\"children\":\"Page Not Found\"}],[\"$\",\"p\",null,{\"className\":\"nf-desc\",\"children\":\"The skill or page you're looking for doesn't exist or has been moved.\"}],[\"$\",\"$L12\",null,{\"className\":\"nf-link\",\"href\":\"/\",\"children\":\"Back to BytesAgain\"}]]}]]}],[]],\"forbidden\":\"$undefined\",\"unauthorized\":\"$undefined\"}]}]\n"])</script><script>self.__next_f.push([1,"7:[\"$\",\"$L13\",null,{}]\n8:[\"$\",\"$L14\",null,{}]\n9:[\"$\",\"$1\",\"c\",{\"children\":[null,[\"$\",\"$L10\",null,{\"parallelRouterKey\":\"children\",\"error\":\"$undefined\",\"errorStyles\":\"$undefined\",\"errorScripts\":\"$undefined\",\"template\":[\"$\",\"$L11\",null,{}],\"templateStyles\":\"$undefined\",\"templateScripts\":\"$undefined\",\"notFound\":\"$undefined\",\"forbidden\":\"$undefined\",\"unauthorized\":\"$undefined\"}]]}]\na:[\"$\",\"$1\",\"c\",{\"children\":[null,[\"$\",\"$L10\",null,{\"parallelRouterKey\":\"children\",\"error\":\"$undefined\",\"errorStyles\":\"$undefined\",\"errorScripts\":\"$undefined\",\"template\":[\"$\",\"$L11\",null,{}],\"templateStyles\":\"$undefined\",\"templateScripts\":\"$undefined\",\"notFound\":\"$undefined\",\"forbidden\":\"$undefined\",\"unauthorized\":\"$undefined\"}]]}]\nb:[\"$\",\"$1\",\"c\",{\"children\":[\"$L15\",[[\"$\",\"script\",\"script-0\",{\"src\":\"/_next/static/chunks/12w5ognupk9fb.js?dpl=dpl_9zPFH6pojPkqyspdDNz28GzCc6KQ\",\"async\":true,\"nonce\":\"$undefined\"}]],[\"$\",\"$L16\",null,{\"children\":[\"$\",\"$17\",null,{\"name\":\"Next.MetadataOutlet\",\"children\":\"$@18\"}]}]]}]\n19:[]\nc:\"$W19\"\nd:[\"$\",\"$1\",\"h\",{\"children\":[null,[\"$\",\"$L1a\",null,{\"children\":\"$L1b\"}],[\"$\",\"div\",null,{\"hidden\":true,\"children\":[\"$\",\"$L1c\",null,{\"children\":[\"$\",\"$17\",null,{\"name\":\"Next.Metadata\",\"children\":\"$L1d\"}]}]}],[\"$\",\"meta\",null,{\"name\":\"next-size-adjust\",\"content\":\"\"}]]}]\nf:[\"$\",\"link\",\"0\",{\"rel\":\"stylesheet\",\"href\":\"/_next/static/chunks/051nc0vy_6.rl.css?dpl=dpl_9zPFH6pojPkqyspdDNz28GzCc6KQ\",\"precedence\":\"next\",\"crossOrigin\":\"$undefined\",\"nonce\":\"$undefined\"}]\n"])</script><script>self.__next_f.push([1,"1b:[[\"$\",\"meta\",\"0\",{\"charSet\":\"utf-8\"}],[\"$\",\"meta\",\"1\",{\"name\":\"viewport\",\"content\":\"width=device-width, initial-scale=1\"}]]\n"])</script><script>self.__next_f.push([1,"1e:I[27201,[\"/_next/static/chunks/0ph_0rx9ah5dc.js?dpl=dpl_9zPFH6pojPkqyspdDNz28GzCc6KQ\",\"/_next/static/chunks/0i_x3w546rsb3.js?dpl=dpl_9zPFH6pojPkqyspdDNz28GzCc6KQ\",\"/_next/static/chunks/06ig5gym-0n-u.js?dpl=dpl_9zPFH6pojPkqyspdDNz28GzCc6KQ\"],\"IconMark\"]\n18:null\n"])</script><script>self.__next_f.push([1,"1d:[[\"$\",\"title\",\"0\",{\"children\":\"SOTA Agent β€” AI Agent Skill | BytesAgain | BytesAgain\"}],[\"$\",\"meta\",\"1\",{\"name\":\"description\",\"content\":\"SOTA Agent is a public ClawHub SOTA-campaign skill for CV and DS work. Use it when the user says \\\"sota agent\\\", \\\"state of the art benchmark scouting\\\", or want...\"}],[\"$\",\"meta\",\"2\",{\"name\":\"robots\",\"content\":\"index, follow\"}],[\"$\",\"meta\",\"3\",{\"name\":\"googlebot\",\"content\":\"index, follow, max-image-preview:large, max-snippet:-1\"}],[\"$\",\"meta\",\"4\",{\"name\":\"llms-txt\",\"content\":\"https://bytesagain.com/llms.txt\"}],[\"$\",\"meta\",\"5\",{\"name\":\"llms-full-txt\",\"content\":\"https://bytesagain.com/llms-full.txt\"}],[\"$\",\"link\",\"6\",{\"rel\":\"canonical\",\"href\":\"https://bytesagain.com/skill/sota-agent\"}],[\"$\",\"meta\",\"7\",{\"name\":\"baidu-site-verification\",\"content\":\"codeva-0evUqX1TFs\"}],[\"$\",\"meta\",\"8\",{\"property\":\"og:title\",\"content\":\"SOTA Agent β€” AI Agent Skill | BytesAgain\"}],[\"$\",\"meta\",\"9\",{\"property\":\"og:description\",\"content\":\"SOTA Agent is a public ClawHub SOTA-campaign skill for CV and DS work. Use it when the user says \\\"sota agent\\\", \\\"state of the art benchmark scouting\\\", or want...\"}],[\"$\",\"meta\",\"10\",{\"property\":\"og:url\",\"content\":\"https://bytesagain.com/skill/sota-agent\"}],[\"$\",\"meta\",\"11\",{\"property\":\"og:site_name\",\"content\":\"BytesAgain\"}],[\"$\",\"meta\",\"12\",{\"property\":\"og:image\",\"content\":\"https://bytesagain.com/social-preview.png\"}],[\"$\",\"meta\",\"13\",{\"property\":\"og:image:width\",\"content\":\"1200\"}],[\"$\",\"meta\",\"14\",{\"property\":\"og:image:height\",\"content\":\"630\"}],[\"$\",\"meta\",\"15\",{\"property\":\"og:type\",\"content\":\"website\"}],[\"$\",\"meta\",\"16\",{\"name\":\"twitter:card\",\"content\":\"summary_large_image\"}],[\"$\",\"meta\",\"17\",{\"name\":\"twitter:title\",\"content\":\"SOTA Agent β€” AI Agent Skill | BytesAgain\"}],[\"$\",\"meta\",\"18\",{\"name\":\"twitter:description\",\"content\":\"SOTA Agent is a public ClawHub SOTA-campaign skill for CV and DS work. Use it when the user says \\\"sota agent\\\", \\\"state of the art benchmark scouting\\\", or want...\"}],[\"$\",\"meta\",\"19\",{\"name\":\"twitter:image\",\"content\":\"https://bytesagain.com/social-preview.png\"}],[\"$\",\"meta\",\"20\",{\"name\":\"twitter:image:width\",\"content\":\"1200\"}],[\"$\",\"meta\",\"21\",{\"name\":\"twitter:image:height\",\"content\":\"630\"}],[\"$\",\"link\",\"22\",{\"rel\":\"icon\",\"href\":\"/favicon.ico?favicon.0x3dzn~oxb6tn.ico\",\"sizes\":\"256x256\",\"type\":\"image/x-icon\"}],[\"$\",\"$L1e\",\"23\",{}]]\n"])</script><script>self.__next_f.push([1,"1f:T1562,"])</script><script>self.__next_f.push([1,"\n .skill-page { max-width: 1100px; margin: 0 auto; padding: 32px 20px 80px; }\n .two-col { display: flex; gap: 32px; align-items: flex-start; }\n .two-col-main { flex: 1; min-width: 0; }\n .two-col-side { width: 300px; flex-shrink: 0; }\n @media (max-width: 860px) {\n .two-col { flex-direction: column; }\n .two-col-side { width: 100%; }\n }\n .breadcrumb { font-size: .82em; color: var(--text-muted2); margin-bottom: 28px; }\n .breadcrumb a { color: #818cf8; text-decoration: none; }\n .breadcrumb a:hover { text-decoration: underline; }\n .skill-card { background: var(--bg-card); border: 1px solid var(--border-card); border-radius: 20px; padding: 28px; margin-bottom: 24px; }\n .skill-header { display: flex; align-items: flex-start; justify-content: space-between; gap: 16px; margin-bottom: 20px; flex-wrap: wrap; }\n .skill-badges { display: flex; align-items: center; gap: 8px; flex-wrap: wrap; }\n .skill-top-actions { display: flex; align-items: center; gap: 10px; margin-left: auto; }\n .badge { display: inline-flex; align-items: center; gap: 5px; font-size: .75em; font-weight: 600; padding: 4px 12px; border-radius: 999px; border: 1px solid transparent; }\n .skill-title { font-size: 1.6em; font-weight: 800; color: var(--text-primary); margin: 0 0 4px; line-height: 1.2; }\n .skill-owner { font-size: .82em; color: var(--text-muted2); margin: 0 0 14px; }\n .skill-owner span { color: #818cf8; }\n .skill-desc { font-size: .92em; color: var(--text-secondary); line-height: 1.65; margin: 0 0 16px; }\n .skill-meta { display: flex; gap: 16px; flex-wrap: wrap; margin-bottom: 18px; padding-bottom: 16px; border-bottom: 1px solid var(--border-card); }\n .meta-item { display: flex; flex-direction: column; gap: 2px; }\n .meta-label { font-size: .7em; color: var(--text-muted5); text-transform: uppercase; letter-spacing: 1px; font-weight: 600; }\n .meta-value { font-size: .92em; color: var(--text-muted2); font-weight: 600; }\n .tags-row { display: flex; gap: 6px; flex-wrap: wrap; }\n .tag { font-size: .75em; color: #6366f1; background: #6366f115; border: 1px solid #6366f130; border-radius: 6px; padding: 3px 10px; text-decoration: none; }\n .tag:hover { background: #6366f125; }\n .install-box { background: var(--bg-deep); border: 1px solid var(--border-card); border-radius: 12px; overflow: hidden; margin-bottom: 24px; }\n .install-header { display: flex; align-items: center; justify-content: space-between; padding: 10px 16px; border-bottom: 1px solid var(--border-card); }\n .install-dots { display: flex; gap: 6px; }\n .dot { width: 10px; height: 10px; border-radius: 50%; }\n .install-label { font-size: .72em; color: var(--text-muted5); font-family: monospace; letter-spacing: 1px; }\n .install-body { padding: 16px 20px; display: flex; align-items: center; justify-content: space-between; gap: 12px; }\n .install-cmd { color: var(--text-code);\n font-family: 'Courier New', monospace; font-size: 1em; }\n .copy-btn { font-size: .75em; color: #6366f1; background: #6366f115; border: 1px solid #6366f130; border-radius: 6px; padding: 5px 12px; cursor: pointer; white-space: nowrap; transition: all .15s; }\n .copy-btn:hover { background: #6366f125; }\n .btn-secondary { display: inline-flex; align-items: center; gap: 8px; padding: 13px 24px; background: transparent; border: 1px solid var(--border-card); border-radius: 10px; color: #6b7280; text-decoration: none; font-weight: 600; font-size: .95em; transition: all .15s; }\n .btn-secondary:hover { border-color: #818cf8; color: #818cf8; }\n .ours-badge { display: inline-flex; align-items: center; gap: 6px; font-size: .72em; font-weight: 700; color: #22d3ee; background: #22d3ee10; border: 1px solid #22d3ee30; border-radius: 999px; padding: 4px 14px; }\n .section-card { background: var(--bg-card); border: 1px solid var(--border-card); border-radius: 16px; padding: 22px 24px; margin-bottom: 20px; }\n .section-title { color: var(--text-primary); font-size: 1.08em; font-weight: 800; margin: 0 0 12px; display: flex; align-items: center; gap: 8px; }\n /* Script box */\n .script-header { display: flex; align-items: center; justify-content: space-between; padding: 8px 14px; background: var(--bg-input); border-bottom: 1px solid var(--border-card); }\n .script-filename { font-size: .72em; color: var(--text-muted2); font-family: 'Courier New', monospace; }\n .script-copy-btn { font-size: .72em; color: #6366f1; background: none; border: 1px solid #6366f130; border-radius: 4px; padding: 2px 10px; cursor: pointer; }\n .script-copy-btn:hover { background: #6366f115; }\n .script-body { padding: 14px 16px; font-family: 'Courier New', monospace; font-size: .82em; line-height: 1.6; color: var(--text-code); overflow-x: auto; max-height: 420px; overflow-y: auto; white-space: pre; }\n /* Articles */\n .article-card { display: block; background: var(--bg-secondary); border: 1px solid var(--border-primary); border-radius: 10px; padding: 14px 16px; text-decoration: none; transition: border-color .15s; }\n .article-card:hover { border-color: #6366f1; }\n @media (max-width: 600px) {\n .skill-card { padding: 20px; }\n .skill-title { font-size: 1.5em; }\n }\n "])</script><script>self.__next_f.push([1,"15:[[\"$\",\"style\",null,{\"children\":\"$1f\"}],\"$L20\",\"$L21\"]\n"])</script><script>self.__next_f.push([1,"22:I[78297,[\"/_next/static/chunks/0ph_0rx9ah5dc.js?dpl=dpl_9zPFH6pojPkqyspdDNz28GzCc6KQ\",\"/_next/static/chunks/0i_x3w546rsb3.js?dpl=dpl_9zPFH6pojPkqyspdDNz28GzCc6KQ\",\"/_next/static/chunks/06ig5gym-0n-u.js?dpl=dpl_9zPFH6pojPkqyspdDNz28GzCc6KQ\",\"/_next/static/chunks/12w5ognupk9fb.js?dpl=dpl_9zPFH6pojPkqyspdDNz28GzCc6KQ\"],\"default\"]\n23:T5808,"])</script><script>self.__next_f.push([1,"\u003cp style=\"margin:8px 0\"\u003e\u003chr style=\"border:none;border-top:1px solid #1e1e3f;margin:12px 0\"\u003e\nname: sota-agent\ndescription: SOTA Agent is a public ClawHub SOTA-campaign skill for CV and DS work. Use it when the user says \"sota agent\", \"state of the art benchmark scouting\", or wants benchmark planning, paper triage, ablation design, and claim review for CV or data-science campaigns.\nversion: 1.4.4\nhomepage: https://zack-dev-cm.github.io/\nuser-invocable: true\nmetadata: {\"openclaw\":{\"homepage\":\"https://zack-dev-cm.github.io/\",\"skillKey\":\"sota-agent\",\"requires\":{\"anyBins\":[\"python3\",\"python\"]}}}\n\u003chr style=\"border:none;border-top:1px solid #1e1e3f;margin:12px 0\"\u003e\u003c/p\u003e\u003cp style=\"margin:8px 0\"\u003e\u003ch2 style=\"color:#f3f4f6;margin:20px 0 10px;font-size:1.15em\"\u003eSOTA Agent\u003c/h2\u003e\u003c/p\u003e\u003cp style=\"margin:8px 0\"\u003eSearch intent: \u003ccode style=\"background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em\"\u003esota agent\u003c/code\u003e, \u003ccode style=\"background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em\"\u003estate of the art benchmark scouting\u003c/code\u003e, \u003ccode style=\"background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em\"\u003ecv benchmark campaign\u003c/code\u003e, \u003ccode style=\"background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em\"\u003egpu vm research workflow\u003c/code\u003e\u003c/p\u003e\u003cp style=\"margin:8px 0\"\u003e\u003ch3 style=\"color:#e5e7eb;margin:18px 0 8px;font-size:1.05em\"\u003eGoal\u003c/h3\u003e\u003c/p\u003e\u003cp style=\"margin:8px 0\"\u003eTurn a vague \"beat the benchmark\" request into a disciplined campaign:\u003c/p\u003e\u003cp style=\"margin:8px 0\"\u003e\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003efixed target metric and split\u003c/li\u003e\n\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003eexplicit literature and leaderboard snapshot\u003c/li\u003e\n\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003ebounded reproduction plan\u003c/li\u003e\n\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003eexplicit handoff to the separate execution lane when runs need external tools\u003c/li\u003e\n\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003eevidence requirements that can be reviewed without relying on live session state\u003c/li\u003e\n\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003eablations that answer one question at a time\u003c/li\u003e\n\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003epromotion only when the claim survives review\u003c/li\u003e\u003c/p\u003e\u003cp style=\"margin:8px 0\"\u003eThis skill is the frontier-planning and candidate-selection layer.\nFor execution artifacts or promotion evidence, pair it with\n\u003ccode style=\"background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em\"\u003edata-science-cv-repro-lab\u003c/code\u003e; this skill stays focused on planning and claim review.\u003c/p\u003e\u003cp style=\"margin:8px 0\"\u003e\u003ch3 style=\"color:#e5e7eb;margin:18px 0 8px;font-size:1.05em\"\u003eUse This Skill When\u003c/h3\u003e\u003c/p\u003e\u003cp style=\"margin:8px 0\"\u003e\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003ethe user wants a CV or DS system pushed toward state-of-the-art results\u003c/li\u003e\n\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003ethe task involves reproducing or surpassing recent papers\u003c/li\u003e\n\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003ethe workflow needs paper triage, leaderboard tracking, or claim review\u003c/li\u003e\n\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003ethe workflow needs a clean handoff to an execution skill after the benchmark contract is frozen\u003c/li\u003e\n\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003ethe user needs experiment management across local runs, notebooks, and long-running jobs\u003c/li\u003e\n\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003ethe question is whether execution evidence supports a SOTA candidate\u003c/li\u003e\n\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003ethe question is whether a candidate is a real SOTA step or only noise, leakage, or benchmark overfitting\u003c/li\u003e\u003c/p\u003e\u003cp style=\"margin:8px 0\"\u003eIf the campaign includes serious execution or release review, use this skill to choose and rank candidates,\nthen use \u003ccode style=\"background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em\"\u003edata-science-cv-repro-lab\u003c/code\u003e as the execution lane.\u003c/p\u003e\u003cp style=\"margin:8px 0\"\u003e\u003ch3 style=\"color:#e5e7eb;margin:18px 0 8px;font-size:1.05em\"\u003eQuick Start\u003c/h3\u003e\u003c/p\u003e\u003cp style=\"margin:8px 0\"\u003e1. Freeze the claim target before touching recipes.\n - Name the task, dataset, metric, split, and target score.\n - Name the current trusted baseline.\n - Name the claim threshold for \"match\", \"beat\", or \"not enough\".\u003c/p\u003e\u003cp style=\"margin:8px 0\"\u003e2. Initialize the campaign records immediately.\n - Use \u003ccode style=\"background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em\"\u003epython3 {baseDir}/scripts/init_sota_campaign.py --root \u003cdir\u003e --campaign-id \u003cid\u003e --title \u003ctitle\u003e\u003c/code\u003e.\n - Use \u003ccode style=\"background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em\"\u003epython3 {baseDir}/scripts/init_sota_leaderboard_snapshot.py --out \u003cjson\u003e --task \u003ctask\u003e --dataset \u003cdataset\u003e --metric \u003cmetric\u003e --split \u003csplit\u003e\u003c/code\u003e.\n - Use \u003ccode style=\"background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em\"\u003epython3 {baseDir}/scripts/init_sota_paper_triage.py --out \u003cjson\u003e --campaign-id \u003cid\u003e --task \u003ctask\u003e\u003c/code\u003e.\n - Use \u003ccode style=\"background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em\"\u003epython3 {baseDir}/scripts/init_sota_program.py --out \u003cjson\u003e --campaign-id \u003cid\u003e --task \u003ctask\u003e --dataset \u003cdataset\u003e --metric \u003cmetric\u003e --split \u003csplit\u003e\u003c/code\u003e when you need one machine-readable benchmark, rerun, delegation, and auth plan.\n - Use \u003ccode style=\"background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em\"\u003epython3 {baseDir}/scripts/init_sota_candidate_card.py --out \u003cjson\u003e --candidate-id \u003cid\u003e --campaign-id \u003cid\u003e --objective \u003cgoal\u003e\u003c/code\u003e.\n - If execution review depends on synced QA runs, runtime sweeps, or benchmark panels, store the paired \u003ccode style=\"background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em\"\u003edata-science-cv-repro-lab\u003c/code\u003e review dashboard path in the program and candidate records before the claim review starts.\n - If external execution evidence exists, record the reviewed artifact manifest path in the program and candidate records instead of acting through a live session.\n - If the review surface needs manual or visual QA, use \u003ccode style=\"background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em\"\u003epython3 {baseDir}/scripts/init_sota_validation_scorecard.py --out \u003cjson\u003e --scorecard-id \u003cid\u003e --surface \u003csurface\u003e\u003c/code\u003e.\n - If an external export bundle matters, use \u003ccode style=\"background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em\"\u003epython3 {baseDir}/scripts/init_sota_artifact_manifest.py --out \u003cjson\u003e --bundle-root \u003cdir\u003e\u003c/code\u003e.\n - If a long execution run is involved, record only sanitized summaries and artifact references in the SOTA campaign files.\u003c/p\u003e\u003cp style=\"margin:8px 0\"\u003e3. Separate the campaign roles even if one agent performs all of them.\n - Scout: papers, leaderboards, repos, and benchmark rules.\n - Reproducer: baseline and top-paper reproduction.\n - Ablator: controlled change sets and compute allocation.\n - Reviewer: contamination, metric drift, and claim integrity.\n - Promoter: final claim or hold decision.\n - Keep the benchmark definition and final claim wording fixed.\n - Use bounded scouting and review lanes for literature triage, repo inspection, per-paper extraction, and hard-case review.\n - For repeated audits, batch over a manifest or CSV instead of free-form context accumulation.\u003c/p\u003e\u003cp style=\"margin:8px 0\"\u003e4. Pick the execution lane explicitly.\n - Execution handoff lane: use \u003ccode style=\"background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em\"\u003edata-science-cv-repro-lab\u003c/code\u003e for external runs and artifact capture.\n - Local lane: cheap falsification, tiny reruns, and artifact review.\u003c/p\u003e\u003cp style=\"margin:8px 0\"\u003e5. Keep file writes inside one campaign workspace.\n - Create one dedicated campaign root and keep every \u003ccode style=\"background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em\"\u003e--out\u003c/code\u003e, \u003ccode style=\"background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em\"\u003e--bundle-root\u003c/code\u003e, and \u003ccode style=\"background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em\"\u003e--output-root\u003c/code\u003e path under it.\n - Do not point the bundled scripts at unrelated home-directory or system paths.\n - Treat \u003ccode style=\"background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em\"\u003escripts/sota_public_safety.py\u003c/code\u003e as the canonical public-redaction layer for URLs, refs, and paths.\u003c/p\u003e\u003cp style=\"margin:8px 0\"\u003e6. Work the SOTA ladder in order.\n - Freeze the benchmark definition and auth rule before using more compute.\n - Reproduce the trusted baseline first.\n - Reproduce one relevant reference result or a close public checkpoint.\n - Build a hypothesis backlog from literature gaps, not vibes.\n - Run narrow ablations before broad recipe churn.\n - Stress the best candidate on the fixed review surfaces.\u003c/p\u003e\u003cp style=\"margin:8px 0\"\u003e7. Claim only on full-surface wins.\n - Fixed benchmark score\n - Reproduced baseline delta\n - Compute or cost context\n - Browser or GUI evidence if that lane mattered\n - Failure-case review\n - Exact evidence bundle\n - Render the final review with \u003ccode style=\"background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em\"\u003epython3 {baseDir}/scripts/render_sota_claim_summary.py --candidate-card \u003cjson\u003e --out \u003cmd\u003e\u003c/code\u003e.\u003c/p\u003e\u003cp style=\"margin:8px 0\"\u003e\u003ch3 style=\"color:#e5e7eb;margin:18px 0 8px;font-size:1.05em\"\u003eOperating Rules\u003c/h3\u003e\u003c/p\u003e\u003cp style=\"margin:8px 0\"\u003e\u003ch4 style=\"color:#d1d5db;margin:14px 0 6px;font-size:.95em\"\u003eCampaign rules\u003c/h4\u003e\u003c/p\u003e\u003cp style=\"margin:8px 0\"\u003e\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003eOne campaign has one target benchmark contract.\u003c/li\u003e\n\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003eDo not let the target metric or split drift midstream.\u003c/li\u003e\n\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003eKeep a short hypothesis backlog and kill low-information ideas quickly.\u003c/li\u003e\n\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003eRecord why each experiment exists before running it.\u003c/li\u003e\u003c/p\u003e\u003cp style=\"margin:8px 0\"\u003e\u003ch4 style=\"color:#d1d5db;margin:14px 0 6px;font-size:.95em\"\u003eCodex multi-agent rules\u003c/h4\u003e\u003c/p\u003e\u003cp style=\"margin:8px 0\"\u003e\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003eMain thread owns the benchmark contract, stop conditions, and final claim decision.\u003c/li\u003e\n\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003eSubagents should do bounded work only: scout, reproduce, ablate, or review.\u003c/li\u003e\n\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003eDo not let one exploratory thread silently rewrite the campaign contract.\u003c/li\u003e\n\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003eFor repeated claim checks or literature extraction, prefer manifest-driven fanout over conversational drift.\u003c/li\u003e\u003c/p\u003e\u003cp style=\"margin:8px 0\"\u003e\u003ch4 style=\"color:#d1d5db;margin:14px 0 6px;font-size:.95em\"\u003eLiterature rules\u003c/h4\u003e\u003c/p\u003e\u003cp style=\"margin:8px 0\"\u003e\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003eRead only the papers or repos that change the candidate plan.\u003c/li\u003e\n\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003eExtract the minimum useful fields: task, metric, split, data, compute, architecture, augmentations, training tricks, and caveats.\u003c/li\u003e\n\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003ePrefer a reproduced strong baseline over copying five tricks from five papers without control.\u003c/li\u003e\n\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003eDo not treat leaderboard rows as ground truth without checking task definition and split rules.\u003c/li\u003e\u003c/p\u003e\u003cp style=\"margin:8px 0\"\u003e\u003ch4 style=\"color:#d1d5db;margin:14px 0 6px;font-size:.95em\"\u003eAblation rules\u003c/h4\u003e\u003c/p\u003e\u003cp style=\"margin:8px 0\"\u003e\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003eChange one meaningful variable at a time when the goal is causal understanding.\u003c/li\u003e\n\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003eIf several knobs move together, label the run as a package change, not an ablation.\u003c/li\u003e\n\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003eKeep one canonical baseline recipe alive for comparison.\u003c/li\u003e\n\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003eRequire the first winning candidate to survive at least one rerun or adjacent-seed check before escalating the claim.\u003c/li\u003e\u003c/p\u003e\u003cp style=\"margin:8px 0\"\u003e\u003ch4 style=\"color:#d1d5db;margin:14px 0 6px;font-size:.95em\"\u003eCompute rules\u003c/h4\u003e\u003c/p\u003e\u003cp style=\"margin:8px 0\"\u003e\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003eSpend cheap compute on reproduction and short falsification first.\u003c/li\u003e\n\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003eDo not push a long run unless the hypothesis would matter if it wins.\u003c/li\u003e\n\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003eRecord training cost, wall time, and hardware for every serious candidate.\u003c/li\u003e\n\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003eCut branches that cannot plausibly clear the target with the remaining budget.\u003c/li\u003e\u003c/p\u003e\u003cp style=\"margin:8px 0\"\u003e\u003ch4 style=\"color:#d1d5db;margin:14px 0 6px;font-size:.95em\"\u003eRuntime and auth rules\u003c/h4\u003e\u003c/p\u003e\u003cp style=\"margin:8px 0\"\u003e\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003eThis public skill does not require API keys, account tokens, live sessions, or account-bound credentials.\u003c/li\u003e\n\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003ePrefer local files, public URLs, and user-supplied artifacts over account-bound execution paths.\u003c/li\u003e\n\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003eDo not require or recommend \u003ccode style=\"background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em\"\u003eOPENAI_API_KEY\u003c/code\u003e, other vendor API keys, or paid inference APIs as the default campaign runtime path.\u003c/li\u003e\n\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003eIf a third-party framework only works through paid API keys, treat it as reference material unless it can run through local tools or public artifacts.\u003c/li\u003e\u003c/p\u003e\u003cp style=\"margin:8px 0\"\u003e\u003ch4 style=\"color:#d1d5db;margin:14px 0 6px;font-size:.95em\"\u003eExternal execution rules\u003c/h4\u003e\u003c/p\u003e\u003cp style=\"margin:8px 0\"\u003e\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003eExecution rules live in \u003ccode style=\"background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em\"\u003edata-science-cv-repro-lab\u003c/code\u003e.\u003c/li\u003e\n\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003eIn this skill, record only the benchmark contract, candidate rationale, review status, and sanitized artifact references.\u003c/li\u003e\u003c/p\u003e\u003cp style=\"margin:8px 0\"\u003e\u003ch4 style=\"color:#d1d5db;margin:14px 0 6px;font-size:.95em\"\u003eClaim safety rules\u003c/h4\u003e\u003c/p\u003e\u003cp style=\"margin:8px 0\"\u003e\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003eNo SOTA claim without a fixed metric, split, and baseline.\u003c/li\u003e\n\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003eNo SOTA claim on a contaminated benchmark or hidden train-on-test path.\u003c/li\u003e\n\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003eIf the execution story depends on a dashboard or synced review surface, keep the dashboard path, source audit, and leakage audit in the claim packet.\u003c/li\u003e\n\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003eIf a candidate wins only on one slice while regressing important surfaces, hold it.\u003c/li\u003e\n\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003eReport uncertainty honestly: \"best internal result so far\" is not the same as \"new SOTA\".\u003c/li\u003e\n\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003eSmall deltas need rerun or adjacent-seed support before they become claim language.\u003c/li\u003e\u003c/p\u003e\u003cp style=\"margin:8px 0\"\u003e\u003ch3 style=\"color:#e5e7eb;margin:18px 0 8px;font-size:1.05em\"\u003eReferences\u003c/h3\u003e\u003c/p\u003e\u003cp style=\"margin:8px 0\"\u003eRead only the reference that matches the task:\u003c/p\u003e\u003cp style=\"margin:8px 0\"\u003e\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003e\u003ccode style=\"background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em\"\u003ereferences/sota-campaign-playbook.md\u003c/code\u003e\u003c/li\u003e\n - Full campaign structure, role separation, and stop conditions.\n\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003e\u003ccode style=\"background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em\"\u003ereferences/sota-program-rules.md\u003c/code\u003e\u003c/li\u003e\n - Rules for queues, stage discipline, ablations, and promotion gating.\n\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003e\u003ccode style=\"background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em\"\u003ereferences/campaign-harness-stack.md\u003c/code\u003e\u003c/li\u003e\n - What to reuse from Codex subagents, harness engineering, OpenEvolve, Symphony, Paperclip, and OptiLLM under a local-first campaign rule.\n\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003e\u003ccode style=\"background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em\"\u003ereferences/benchmark-discipline.md\u003c/code\u003e\u003c/li\u003e\n - How to avoid contamination, metric drift, and invalid comparisons.\n\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003e\u003ccode style=\"background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em\"\u003ereferences/paper-triage.md\u003c/code\u003e\u003c/li\u003e\n - How to filter papers and extract only decision-relevant details.\n\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003e\u003ccode style=\"background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em\"\u003ereferences/public-research-lane.md\u003c/code\u003e\u003c/li\u003e\n - How to review public literature and leaderboard pages without private sessions.\n\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003e\u003ccode style=\"background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em\"\u003ereferences/external-evidence-handoff.md\u003c/code\u003e\u003c/li\u003e\n - How to record sanitized evidence from external notebook or UI runs without controlling a live session.\n\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003e\u003ccode style=\"background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em\"\u003ereferences/execution-evidence-summary.md\u003c/code\u003e\u003c/li\u003e\n - How to summarize execution evidence that belongs in the paired execution skill.\n\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003e\u003ccode style=\"background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em\"\u003ereferences/claim-safety.md\u003c/code\u003e\u003c/li\u003e\n - Review rules for whether a candidate deserves a SOTA claim at all.\n\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003e\u003ccode style=\"background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em\"\u003ereferences/public-safety.md\u003c/code\u003e\u003c/li\u003e\n - Publication review rules for secrets, private refs, and raw notebook paths.\u003c/p\u003e\u003cp style=\"margin:8px 0\"\u003e\u003ch3 style=\"color:#e5e7eb;margin:18px 0 8px;font-size:1.05em\"\u003eBundled Scripts\u003c/h3\u003e\u003c/p\u003e\u003cp style=\"margin:8px 0\"\u003e\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003e\u003ccode style=\"background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em\"\u003escripts/sota_public_safety.py\u003c/code\u003e\u003c/li\u003e\n - Pure local helpers for path, URL, ref, env, and command redaction. No network I/O or subprocess execution.\n\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003e\u003ccode style=\"background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em\"\u003escripts/init_sota_campaign.py\u003c/code\u003e\u003c/li\u003e\n - Create a reusable campaign folder with benchmark, program, agent, research, leaderboard, plan, ablation, evidence, and claim files.\n\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003e\u003ccode style=\"background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em\"\u003escripts/init_sota_program.py\u003c/code\u003e\u003c/li\u003e\n - Create a machine-readable program record with the fixed benchmark, baselines, rerun policy, bounded subagent roles, and local-first runtime rules.\n\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003e\u003ccode style=\"background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em\"\u003escripts/init_sota_leaderboard_snapshot.py\u003c/code\u003e\u003c/li\u003e\n - Create a machine-readable snapshot of the target benchmark contract and current reference scores.\n\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003e\u003ccode style=\"background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em\"\u003escripts/init_sota_paper_triage.py\u003c/code\u003e\u003c/li\u003e\n - Create a machine-readable literature queue for paper screening and extraction.\n\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003e\u003ccode style=\"background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em\"\u003escripts/init_sota_browser_run_card.py\u003c/code\u003e\u003c/li\u003e\n - Create a sanitized external-evidence record for notebook or UI run artifacts.\n\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003e\u003ccode style=\"background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em\"\u003escripts/init_sota_validation_scorecard.py\u003c/code\u003e\u003c/li\u003e\n - Create a machine-readable GUI or notebook validation scorecard when visible state matters to the campaign.\n\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003e\u003ccode style=\"background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em\"\u003escripts/init_sota_artifact_manifest.py\u003c/code\u003e\u003c/li\u003e\n - Create a machine-readable export-bundle manifest for external artifacts with redacted public path metadata.\n\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003e\u003ccode style=\"background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em\"\u003escripts/init_sota_candidate_card.py\u003c/code\u003e\u003c/li\u003e\n - Create a machine-readable card for a serious candidate, its execution lane, auth mode, and claim state.\n\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003e\u003ccode style=\"background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em\"\u003escripts/init_sota_candidate.py\u003c/code\u003e\u003c/li\u003e\n - Create a machine-readable candidate record with change set, risks, and redacted public artifact refs.\n\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003e\u003ccode style=\"background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em\"\u003escripts/init_sota_ablation_queue.py\u003c/code\u003e\u003c/li\u003e\n - Create a focused ablation queue for one candidate family.\n\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003e\u003ccode style=\"background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em\"\u003escripts/init_sota_vm_bootstrap_manifest.py\u003c/code\u003e\u003c/li\u003e\n - Create a redacted long-run summary manifest for already-approved execution artifacts.\n\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003e\u003ccode style=\"background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em\"\u003escripts/update_sota_scoreboard.py\u003c/code\u003e\u003c/li\u003e\n - Refresh a ranked scoreboard for a fixed metric and goal direction.\n\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003e\u003ccode style=\"background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em\"\u003escripts/init_sota_review_packet.py\u003c/code\u003e\u003c/li\u003e\n - Join the core artifacts for a promotion, hold, or cut decision.\n\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003e\u003ccode style=\"background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em\"\u003escripts/render_sota_claim_summary.py\u003c/code\u003e\u003c/li\u003e\n - Render a concise markdown review from the machine-readable candidate card.\n\u003cli style=\"color:#94a3b8;margin:3px 0\"\u003e\u003ccode style=\"background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em\"\u003escripts/render_sota_program_summary.py\u003c/code\u003e\u003c/li\u003e\n - Render a concise markdown summary from the program, candidate, scoreboard, and review packet.\n\u003c/p\u003e"])</script><script>self.__next_f.push([1,"20:[\"$\",\"div\",null,{\"className\":\"skill-page\",\"children\":[[\"$\",\"script\",null,{\"type\":\"application/ld+json\",\"dangerouslySetInnerHTML\":{\"__html\":\"{\\\"@context\\\":\\\"https://schema.org\\\",\\\"@type\\\":\\\"SoftwareApplication\\\",\\\"name\\\":\\\"SOTA Agent\\\",\\\"description\\\":\\\"SOTA Agent is a public ClawHub SOTA-campaign skill for CV and DS work. Use it when the user says \\\\\\\"sota agent\\\\\\\", \\\\\\\"state of the art benchmark scouting\\\\\\\", or want...\\\",\\\"url\\\":\\\"https://bytesagain.com/skill/sota-agent\\\",\\\"applicationCategory\\\":\\\"requires-sensitive-credentials\\\",\\\"operatingSystem\\\":\\\"Any\\\",\\\"offers\\\":{\\\"@type\\\":\\\"Offer\\\",\\\"price\\\":\\\"0\\\",\\\"priceCurrency\\\":\\\"USD\\\"},\\\"publisher\\\":{\\\"@type\\\":\\\"Organization\\\",\\\"name\\\":\\\"BytesAgain\\\",\\\"url\\\":\\\"https://bytesagain.com\\\"}}\"}}],[\"$\",\"div\",null,{\"className\":\"breadcrumb\",\"children\":[[\"$\",\"a\",null,{\"href\":\"/\",\"children\":\"BytesAgain\"}],\" β€Ί \",[\"$\",\"a\",null,{\"href\":\"/skills\",\"children\":\"Skills\"}],\" β€Ί \",\"SOTA Agent\"]}],[\"$\",\"div\",null,{\"className\":\"two-col\",\"children\":[[\"$\",\"div\",null,{\"className\":\"two-col-main\",\"children\":[[\"$\",\"div\",null,{\"className\":\"skill-card\",\"children\":[[\"$\",\"div\",null,{\"className\":\"skill-header\",\"children\":[[\"$\",\"div\",null,{\"className\":\"skill-badges\",\"children\":[[\"$\",\"span\",null,{\"className\":\"badge\",\"style\":{\"color\":\"#818cf8\",\"background\":\"#818cf822\",\"borderColor\":\"#818cf844\"},\"children\":[\"πŸ¦€\",\" \",\"ClawHub\"]}],false]}],[\"$\",\"div\",null,{\"className\":\"skill-top-actions\",\"children\":[\"$\",\"$L22\",null,{\"slug\":\"sota-agent\"}]}]]}],[\"$\",\"h1\",null,{\"className\":\"skill-title\",\"children\":\"SOTA Agent\"}],[\"$\",\"p\",null,{\"className\":\"skill-owner\",\"children\":[\"by \",[\"$\",\"span\",null,{\"children\":[\"@\",\"zack-dev-cm\"]}]]}],[\"$\",\"p\",null,{\"className\":\"skill-desc\",\"children\":\"SOTA Agent is a public ClawHub SOTA-campaign skill for CV and DS work. Use it when the user says \\\"sota agent\\\", \\\"state of the art benchmark scouting\\\", or want...\"}],[\"$\",\"div\",null,{\"className\":\"skill-meta\",\"children\":[[\"$\",\"div\",null,{\"className\":\"meta-item\",\"children\":[[\"$\",\"span\",null,{\"className\":\"meta-label\",\"children\":\"Version\"}],[\"$\",\"span\",null,{\"className\":\"meta-value\",\"children\":[\"v\",\"1.4.4\"]}]]}],[\"$\",\"div\",null,{\"className\":\"meta-item\",\"children\":[[\"$\",\"span\",null,{\"className\":\"meta-label\",\"children\":\"Downloads\"}],[\"$\",\"span\",null,{\"className\":\"meta-value\",\"children\":\"939\"}]]}],false,[\"$\",\"div\",null,{\"className\":\"meta-item\",\"children\":[[\"$\",\"span\",null,{\"className\":\"meta-label\",\"children\":\"Stars\"}],[\"$\",\"span\",null,{\"className\":\"meta-value\",\"children\":[\"⭐ \",\"2\"]}]]}],false,[\"$\",\"div\",null,{\"className\":\"meta-item\",\"style\":{\"flexDirection\":\"row\",\"gap\":6,\"alignItems\":\"center\"},\"children\":[[\"$\",\"a\",\"legal\",{\"href\":\"/?q=legal\",\"className\":\"tag\",\"children\":[\"#\",\"legal\"]}],[\"$\",\"a\",\"hr\",{\"href\":\"/?q=hr\",\"className\":\"tag\",\"children\":[\"#\",\"hr\"]}]]}]]}],[\"$\",\"div\",null,{\"style\":{\"marginTop\":6},\"children\":[\"$\",\"a\",null,{\"href\":\"https://clawhub.ai/zack-dev-cm/sota-agent\",\"target\":\"_blank\",\"rel\":\"noopener\",\"className\":\"btn-secondary\",\"style\":{\"padding\":\"6px 12px\",\"fontSize\":\".82em\",\"borderRadius\":8,\"background\":\"transparent\",\"border\":\"1px solid var(--border-card)\",\"color\":\"var(--text-muted2)\",\"textDecoration\":\"none\",\"whiteSpace\":\"nowrap\"},\"children\":[\"View on \",\"ClawHub\",\" β†’\"]}]}]]}],[\"$\",\"div\",null,{\"className\":\"install-box\",\"children\":[[\"$\",\"div\",null,{\"className\":\"install-header\",\"children\":[[\"$\",\"div\",null,{\"className\":\"install-dots\",\"children\":[[\"$\",\"div\",null,{\"className\":\"dot\",\"style\":{\"background\":\"#ef4444\"}}],[\"$\",\"div\",null,{\"className\":\"dot\",\"style\":{\"background\":\"#eab308\"}}],[\"$\",\"div\",null,{\"className\":\"dot\",\"style\":{\"background\":\"#22c55e\"}}]]}],[\"$\",\"span\",null,{\"className\":\"install-label\",\"children\":\"TERMINAL\"}]]}],[\"$\",\"div\",null,{\"className\":\"install-body\",\"style\":{\"flexWrap\":\"wrap\"},\"children\":[[\"$\",\"code\",null,{\"className\":\"install-cmd\",\"children\":\"clawhub install sota-agent\"}],[\"$\",\"button\",null,{\"className\":\"copy-btn\",\"data-cmd\":\"clawhub install sota-agent\",\"style\":{\"fontWeight\":700},\"children\":\"Copy\"}]]}]]}],[\"$\",\"section\",null,{\"className\":\"skill-card\",\"style\":{\"marginBottom\":20},\"children\":[[\"$\",\"h2\",null,{\"style\":{\"color\":\"#f8fafc\",\"fontSize\":\"1.2em\",\"fontWeight\":800,\"margin\":\"0 0 16px\",\"display\":\"flex\",\"alignItems\":\"center\",\"gap\":8},\"children\":\"πŸ“– About This Skill\"}],[\"$\",\"div\",null,{\"style\":{\"fontSize\":\".92em\",\"color\":\"#94a3b8\",\"lineHeight\":1.75},\"dangerouslySetInnerHTML\":{\"__html\":\"$23\"}}]]}],null,\"$L24\",null,null,null,null,null,false,false]}],\"$L25\"]}]]}]\n"])</script><script>self.__next_f.push([1,"21:[\"$\",\"script\",null,{\"dangerouslySetInnerHTML\":{\"__html\":\"\\n document.querySelectorAll('.copy-btn, .script-copy-btn').forEach(btn =\u003e {\\n btn.addEventListener('click', () =\u003e {\\n const cmd = btn.getAttribute('data-cmd');\\n if (!cmd) return;\\n navigator.clipboard.writeText(cmd).then(() =\u003e {\\n const orig = btn.textContent;\\n btn.textContent = 'Copied!';\\n setTimeout(() =\u003e btn.textContent = orig, 1500);\\n }).catch(() =\u003e {});\\n });\\n });\\n \"}}]\n"])</script><script>self.__next_f.push([1,"27:I[71521,[\"/_next/static/chunks/0ph_0rx9ah5dc.js?dpl=dpl_9zPFH6pojPkqyspdDNz28GzCc6KQ\",\"/_next/static/chunks/0i_x3w546rsb3.js?dpl=dpl_9zPFH6pojPkqyspdDNz28GzCc6KQ\",\"/_next/static/chunks/06ig5gym-0n-u.js?dpl=dpl_9zPFH6pojPkqyspdDNz28GzCc6KQ\",\"/_next/static/chunks/12w5ognupk9fb.js?dpl=dpl_9zPFH6pojPkqyspdDNz28GzCc6KQ\"],\"default\"]\n26:T1547,"])</script><script>self.__next_f.push([1,"\u003cp style=\"margin:8px 0\"\u003e1. Freeze the claim target before touching recipes.\n - Name the task, dataset, metric, split, and target score.\n - Name the current trusted baseline.\n - Name the claim threshold for \"match\", \"beat\", or \"not enough\".\u003c/p\u003e\u003cp style=\"margin:8px 0\"\u003e2. Initialize the campaign records immediately.\n - Use \u003ccode style=\"background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em\"\u003epython3 {baseDir}/scripts/init_sota_campaign.py --root \u003cdir\u003e --campaign-id \u003cid\u003e --title \u003ctitle\u003e\u003c/code\u003e.\n - Use \u003ccode style=\"background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em\"\u003epython3 {baseDir}/scripts/init_sota_leaderboard_snapshot.py --out \u003cjson\u003e --task \u003ctask\u003e --dataset \u003cdataset\u003e --metric \u003cmetric\u003e --split \u003csplit\u003e\u003c/code\u003e.\n - Use \u003ccode style=\"background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em\"\u003epython3 {baseDir}/scripts/init_sota_paper_triage.py --out \u003cjson\u003e --campaign-id \u003cid\u003e --task \u003ctask\u003e\u003c/code\u003e.\n - Use \u003ccode style=\"background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em\"\u003epython3 {baseDir}/scripts/init_sota_program.py --out \u003cjson\u003e --campaign-id \u003cid\u003e --task \u003ctask\u003e --dataset \u003cdataset\u003e --metric \u003cmetric\u003e --split \u003csplit\u003e\u003c/code\u003e when you need one machine-readable benchmark, rerun, delegation, and auth plan.\n - Use \u003ccode style=\"background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em\"\u003epython3 {baseDir}/scripts/init_sota_candidate_card.py --out \u003cjson\u003e --candidate-id \u003cid\u003e --campaign-id \u003cid\u003e --objective \u003cgoal\u003e\u003c/code\u003e.\n - If execution review depends on synced QA runs, runtime sweeps, or benchmark panels, store the paired \u003ccode style=\"background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em\"\u003edata-science-cv-repro-lab\u003c/code\u003e review dashboard path in the program and candidate records before the claim review starts.\n - If external execution evidence exists, record the reviewed artifact manifest path in the program and candidate records instead of acting through a live session.\n - If the review surface needs manual or visual QA, use \u003ccode style=\"background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em\"\u003epython3 {baseDir}/scripts/init_sota_validation_scorecard.py --out \u003cjson\u003e --scorecard-id \u003cid\u003e --surface \u003csurface\u003e\u003c/code\u003e.\n - If an external export bundle matters, use \u003ccode style=\"background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em\"\u003epython3 {baseDir}/scripts/init_sota_artifact_manifest.py --out \u003cjson\u003e --bundle-root \u003cdir\u003e\u003c/code\u003e.\n - If a long execution run is involved, record only sanitized summaries and artifact references in the SOTA campaign files.\u003c/p\u003e\u003cp style=\"margin:8px 0\"\u003e3. Separate the campaign roles even if one agent performs all of them.\n - Scout: papers, leaderboards, repos, and benchmark rules.\n - Reproducer: baseline and top-paper reproduction.\n - Ablator: controlled change sets and compute allocation.\n - Reviewer: contamination, metric drift, and claim integrity.\n - Promoter: final claim or hold decision.\n - Keep the benchmark definition and final claim wording fixed.\n - Use bounded scouting and review lanes for literature triage, repo inspection, per-paper extraction, and hard-case review.\n - For repeated audits, batch over a manifest or CSV instead of free-form context accumulation.\u003c/p\u003e\u003cp style=\"margin:8px 0\"\u003e4. Pick the execution lane explicitly.\n - Execution handoff lane: use \u003ccode style=\"background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em\"\u003edata-science-cv-repro-lab\u003c/code\u003e for external runs and artifact capture.\n - Local lane: cheap falsification, tiny reruns, and artifact review.\u003c/p\u003e\u003cp style=\"margin:8px 0\"\u003e5. Keep file writes inside one campaign workspace.\n - Create one dedicated campaign root and keep every \u003ccode style=\"background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em\"\u003e--out\u003c/code\u003e, \u003ccode style=\"background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em\"\u003e--bundle-root\u003c/code\u003e, and \u003ccode style=\"background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em\"\u003e--output-root\u003c/code\u003e path under it.\n - Do not point the bundled scripts at unrelated home-directory or system paths.\n - Treat \u003ccode style=\"background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em\"\u003escripts/sota_public_safety.py\u003c/code\u003e as the canonical public-redaction layer for URLs, refs, and paths.\u003c/p\u003e\u003cp style=\"margin:8px 0\"\u003e6. Work the SOTA ladder in order.\n - Freeze the benchmark definition and auth rule before using more compute.\n - Reproduce the trusted baseline first.\n - Reproduce one relevant reference result or a close public checkpoint.\n - Build a hypothesis backlog from literature gaps, not vibes.\n - Run narrow ablations before broad recipe churn.\n - Stress the best candidate on the fixed review surfaces.\u003c/p\u003e\u003cp style=\"margin:8px 0\"\u003e7. Claim only on full-surface wins.\n - Fixed benchmark score\n - Reproduced baseline delta\n - Compute or cost context\n - Browser or GUI evidence if that lane mattered\n - Failure-case review\n - Exact evidence bundle\n - Render the final review with \u003ccode style=\"background:#0d0d1e;color:#a5f3fc;padding:1px 5px;border-radius:3px;font-size:.88em\"\u003epython3 {baseDir}/scripts/render_sota_claim_summary.py --candidate-card \u003cjson\u003e --out \u003cmd\u003e\u003c/code\u003e.\u003c/p\u003e"])</script><script>self.__next_f.push([1,"24:[\"$\",\"section\",null,{\"className\":\"skill-card\",\"style\":{\"marginBottom\":20},\"children\":[[\"$\",\"h2\",null,{\"style\":{\"color\":\"#f8fafc\",\"fontSize\":\"1.2em\",\"fontWeight\":800,\"margin\":\"0 0 16px\",\"display\":\"flex\",\"alignItems\":\"center\",\"gap\":8},\"children\":\"πŸ’‘ Examples\"}],[\"$\",\"div\",null,{\"style\":{\"fontSize\":\".92em\",\"color\":\"#94a3b8\",\"lineHeight\":1.75},\"dangerouslySetInnerHTML\":{\"__html\":\"$26\"}}]]}]\n25:[\"$\",\"div\",null,{\"className\":\"two-col-side\",\"children\":[\"$\",\"$L27\",null,{\"category\":\"requires-sensitive-credentials\",\"currentSlug\":\"sota-agent\",\"name\":\"SOTA Agent\",\"tags\":[\"legal\",\"hr\"]}]}]\n"])</script></body></html>