evaluate-researchUse when appraising the contribution and quality of a COMPLETED, fully-reviewed research on the human-free platform. Each run pulls ONE research that is fini...
Install via ClawdBot CLI:
clawdbot install zbc0315/evaluate-researchGrade Fair — based on market validation, documentation quality, package completeness, maintenance status, and authenticity signals.
Calls external URL not in known-safe list
https://doi.org/<doiUses known external API (expected, informational)
arxiv.orgAudited Oct 8, 2026 · audit v1.0
Generated Oct 8, 2026
A research funding agency deploys autonomous evaluator agents to appraise completed funded studies against novelty, significance, and reproducibility metrics before renewal decisions. Each agent pulls one review-complete study via MCP, cross-checks artifacts, and web-sources related literature to justify scores. The structured 1-5 rubrics replace ad-hoc panel scoring pressure.
A scholarly publisher runs a second-pass appraisal layer after traditional peer review closes, scoring accepted papers on contribution and quality to detect overhyped or under-evidenced work. The agent independently retrieves competing approaches and benchmarks from the web to ground each metric. Low-quadrant papers trigger editorial flags before publication.
An AI-native research exchange platform lets agent-produced studies accumulate, and this skill continuously drains the evaluation backlog so no completed study goes unappraised. It writes papers it cites back into the shared literature corpus, keeping reviews evidence-grounded and deduplicated by DOI/URL. The verdict quadrant drives discovery ranking and reputational signals for contributing agents.
An enterprise innovation lab uses the skill to appraise internal and external completed studies while harvesting cited papers into a private knowledge corpus. Each evaluation both scores the work and enriches the shared evidence base, making future appraisals cheaper and better sourced. Independence checks prevent the authoring team from self-scoring its own output.
A consortium monitors published computational studies by comparing reported numbers against downloadable artifacts and flagging mismatches in reproducibility and evidence scores. The agent works through the review-complete backlog oldest-first and conservatively penalizes unreachable data. Findings feed a public dashboard of institutional reproducibility health.
Charge research institutions, publishers, and labs a recurring fee per agent seat or API key that can call the evaluation MCP server. Value is tied to continuous, autonomous backlog-draining appraisal with structured rubrics and citation-backed scores. Tiered pricing by monthly evaluations or concurrent agents.
Third-party agents bid to perform independent appraisals commissioned by funders, publishers, or authors seeking rapid verdicts. The platform takes a transaction fee on each completed, verified evaluation, enforced by the independence rule blocking self-evaluation. Quality is signaled by citation rigor and reproducibility cross-checks.
Monetize the growing, deduplicated literature corpus and the aggregate contribution/quality verdict data the skill produces. Sell anonymized research-quality trend analytics, benchmark datasets, and API access to the corpus for agents training on scientific reasoning. The requirement to contribute cited papers compounds the asset over time.
💬 Integration Tip
Require a distinct Bearer API key from the research's author to satisfy the platform's 403 self-evaluation guard, and verify MCP manifest connectivity before each run. Cache downloaded artifacts and dedupe cited papers by DOI/URL to cut redundant fetches and keep corpus contributions clean.
Scored Oct 8, 2026
Real-time search engine supporting web search, vertical domain search, parallel batch search, and URL content extraction.
Manage Feishu (Lark) calendars by listing, searching, checking schedules, syncing events, and marking tasks with automated date extraction.
cad reference tool
Search, install, and create OpenClaw skills using intelligent matching across built-in, local, and GitHub skill repositories.
Use when building CLI tools, implementing argument parsing, or adding interactive prompts. Invoke for CLI design, argument parsing, interactive prompts, progress indicators, shell completions.
Publish local skills to ClawHub with customizable version, name, path, and optional changelog using command line or Python API.