OpenAgentSkill Resolve API Task: scrape pricing pages Policy: human_review_required Recommendation: Best skill: Crawl4AI (crawl4ai) URL: https://www.openagentskill.com/skills/crawl4ai Repository: https://github.com/unclecode/crawl4ai Install: npx skills add unclecode/crawl4ai Install policy: review Trust Score v5: 74/100 Sandbox only Trust decision: Compare alternatives before installing. Risk: Safe to try; Reviewed; 82/100 Strong shortlist Why: - Matches task terms: scrape - Strong GitHub adoption: 66,000 stars - Quality score 96/100 - Install handoff is available - Repository freshness signal is available - Primary pick for Web scraping - 74/100 OpenAgentSkill Trust Score v5 - 89/100 audit score Agent instruction: Use Crawl4AI for: scrape pricing pages Review audit: https://www.openagentskill.com/skills/crawl4ai/audit Ask for human approval before installing: npx skills add unclecode/crawl4ai If the audit or policy warnings are unacceptable, choose one of the alternatives. Agent contract: - Version: openagentskill-resolve-contract-v2 - Install policy: review - Auto-install allowed: no - Human review required: yes - Audit URL: https://www.openagentskill.com/skills/crawl4ai/audit - Do not use when: - teams that need a vendor-supported SLA - high-compliance environments without internal security review - No OpenAgentSkill engagement data yet - Documentation summary is thin Feedback: Event ID: resolve_62500232-3762-49a5-8aa1-b422a89dabf5 Outcome API: https://www.openagentskill.com/api/agent/outcome CLI: npx openagentskill outcome "resolve_62500232-3762-49a5-8aa1-b422a89dabf5" --skill "crawl4ai" --task "scrape pricing pages" --agent "codex" --outcome success Agent Feedback Loop: Version: openagentskill-agent-feedback-loop-v3 Dry run: supported Expected outcomes: success, failed, not_relevant, blocked_by_risk, setup_required Error types: install_failed, runtime_error, not_relevant, permission_blocked, risk_blocked, setup_required, low_quality_output, missing_dependency, timeout, other Workspaces: sandbox, local, ci, production, unknown Updates: Trust Score v5 outcome confidence; Resolve ranking outcome evidence; Skill detail outcome signal block; Outcome leaderboard Install Receipt: OpenAgentSkill Install Receipt Version: openagentskill-install-receipt-v1 Receipt ID: receipt_62500232-3762-49a5-8aa1-b422a89dabf5 Resolve event: resolve_62500232-3762-49a5-8aa1-b422a89dabf5 Generated: 2026-07-28T11:20:45.109Z Task: scrape pricing pages Agent: codex Selected skill: Crawl4AI (crawl4ai) URL: https://www.openagentskill.com/skills/crawl4ai Repository: https://github.com/unclecode/crawl4ai Audit: https://www.openagentskill.com/skills/crawl4ai/audit Eval: https://www.openagentskill.com/api/agent/evals?slug=crawl4ai&task=scrape%20pricing%20pages&max_risk=medium Install: npx skills add unclecode/crawl4ai Install policy: review Auto-install allowed: no Human review required: yes Sandbox first: yes Trust: 74/100 Sandbox only Agent Proven: 0/100 Needs first agent run Audit: 89/100 Safe to try Safety: 77/100 Reviewed Outcome signal: No agent outcome data yet Why selected: - Matches task terms: scrape - Strong GitHub adoption: 66,000 stars - Quality score 96/100 - Install handoff is available - Repository freshness signal is available - Primary pick for Web scraping - 74/100 Trust Score - 0/100 Agent Proven Score Do not use when: - teams that need a vendor-supported SLA - high-compliance environments without internal security review - No OpenAgentSkill engagement data yet - Documentation summary is thin - README/SKILL.md completeness: Public metadata needs stronger README/SKILL.md context - Production credentials, payments, or irreversible account changes without explicit human review - Sensitive private data before reviewing repository code, license, and permission surface - Automatic installation in a production workspace Alternatives: - Firecrawl (firecrawl): npx skills add mendableai/firecrawl - Webapp Testing (anthropic-webapp-testing): npx skills add anthropics/skills --skill webapp-testing - Playwright Browser Skill (openai-playwright): npx skills add openai/skills --skill playwright - Browser Use (browser-use): npx skills add browser-use/browser-use Outcome feedback: Endpoint: POST https://www.openagentskill.com/api/agent/outcome Event ID: resolve_62500232-3762-49a5-8aa1-b422a89dabf5 CLI: npx openagentskill outcome "resolve_62500232-3762-49a5-8aa1-b422a89dabf5" --skill "crawl4ai" --task "scrape pricing pages" --agent "codex" --outcome success Next steps: 1. Read the audit page: https://www.openagentskill.com/skills/crawl4ai/audit 2. Fetch the pre-install eval: https://www.openagentskill.com/api/agent/evals?slug=crawl4ai&task=scrape%20pricing%20pages&max_risk=medium 3. Do not auto-install yet. Ask for human review before running: npx skills add unclecode/crawl4ai 4. Run one narrow task with no production secrets or irreversible side effects. 5. Report the result to https://www.openagentskill.com/api/agent/outcome using event_id resolve_62500232-3762-49a5-8aa1-b422a89dabf5. Decision Packet: Version: openagentskill-agent-decision-packet-v1 Selected: Crawl4AI (crawl4ai) Install: npx skills add unclecode/crawl4ai Policy: review Trust: 74/100 Sandbox only Trust decision: Compare alternatives before installing. Risk: Safe to try; Reviewed Outcome endpoint: https://www.openagentskill.com/api/agent/outcome Outcome event: resolve_62500232-3762-49a5-8aa1-b422a89dabf5 Do not use when: - teams that need a vendor-supported SLA - high-compliance environments without internal security review - No OpenAgentSkill engagement data yet - Documentation summary is thin Alternatives: - Firecrawl: npx skills add mendableai/firecrawl - Webapp Testing: npx skills add anthropics/skills --skill webapp-testing - Playwright Browser Skill: npx skills add openai/skills --skill playwright Selected: Crawl4AI (crawl4ai) Match: 99 Supply: Research / Research agents Maintenance: 2mo since push Risk: Safe to try Safety: 77/100 Review before install Safety tier: Reviewed Auto-install policy: review Audit: 89/100 Safe to try Install: Install the "Crawl4AI" agent skill from https://github.com/unclecode/crawl4ai. Read its SKILL.md or equivalent instructions first, install only the files needed for this workspace, and summarize any required setup before using it. Skill purpose: Open-source LLM-friendly web crawler and scraper for agent workflows. URL: https://www.openagentskill.com/skills/crawl4ai Why: Matches task terms: scrape; Strong GitHub adoption: 66,000 stars; Quality score 96/100; Install handoff is available; Repository freshness signal is available Agent Decision: Recommended skill: Crawl4AI Install command: npx skills add unclecode/crawl4ai Safety gate: Reviewed (review) Why recommended: - Matches task terms: scrape - Strong GitHub adoption: 66,000 stars - Quality score 96/100 - Install handoff is available - Repository freshness signal is available - Primary pick for Web scraping Risk summary: Safe to try; Reviewed; 77/100 Review before install; 82/100 Strong shortlist Risk notes: - Documentation summary is thin - README/SKILL.md completeness: Public metadata needs stronger README/SKILL.md context Next steps: - Read the audit URL before installing. - Ask for human approval before installing. - Use install command: npx skills add unclecode/crawl4ai - After one narrow task, report the outcome with: npx openagentskill outcome "resolve_62500232-3762-49a5-8aa1-b422a89dabf5" --skill "crawl4ai" --task "scrape pricing pages" --agent "codex" --outcome success - Summarize output, warnings, files touched, and whether setup or human review was required. Agent Workflow: Action: Review the audit page, then allow agent install in a sandboxed workflow. Install API: https://www.openagentskill.com/api/skills/crawl4ai/install Command: npx skills add unclecode/crawl4ai Prompt: Task: scrape pricing pages Use Crawl4AI from https://www.openagentskill.com/skills/crawl4ai. Review the audit first: https://www.openagentskill.com/skills/crawl4ai/audit Review the eval first: https://www.openagentskill.com/api/agent/evals?slug=crawl4ai&task=scrape%20pricing%20pages&max_risk=medium Install handoff: https://www.openagentskill.com/api/skills/crawl4ai/install Install command: npx skills add unclecode/crawl4ai If audit or policy warnings look unsafe for this workspace, use one of the alternatives instead. Review checklist: - Safety tier: Reviewed - Safety score: 77/100 Review before install - Audit score: 89/100 Safe to try - Trust score: 74/100 Sandbox only - Readiness: 100/100 Production-ready - Documentation summary is thin - Documentation summary is thin - README/SKILL.md completeness: Public metadata needs stronger README/SKILL.md context Agent Handoff: Version: openagentskill-agent-handoff-v1 Mode: resolve_compare_review_install Selected skill: Crawl4AI (crawl4ai) Install policy: review Auto-install allowed: no Human review required: yes Platform templates: - Codex: use copy_prompt from agent_handoff.platform_templates.codex - Claude Code: use copy_prompt from agent_handoff.platform_templates.claude-code - Cursor: use copy_prompt from agent_handoff.platform_templates.cursor Blocked actions: - Do not install when safety_gate.blocked is true. - Do not install when the audit or eval reports unacceptable workspace risk. - Do not execute shell commands, access secrets, or call external services without user approval. - Do not keep the skill as an always-on instruction if it only fits a narrow task. Alternatives: 2. Firecrawl (firecrawl) — Reviewed, safety 73/100, audit 89/100 3. Webapp Testing (anthropic-webapp-testing) — Reviewed, safety 77/100, audit 93/100 4. Playwright Browser Skill (openai-playwright) — Reviewed with permission notes, safety 63/100, audit 91/100 5. Browser Use (browser-use) — Reviewed with permission notes, safety 70/100, audit 86/100 Blocked for auto-install: No blocked candidates