OpenAgentSkill Agent Integration Kit Version: openagentskill-agent-integration-kit-v2 Purpose: Let an AI agent resolve a task into the right reusable skill, compare alternatives, review trust signals, and install with a safe handoff. Canonical: - Page: https://www.openagentskill.com/agent/integration-kit - API: https://www.openagentskill.com/api/agent/integration-kit - Resolve API: https://www.openagentskill.com/api/agent/resolve - Receipt API: https://www.openagentskill.com/api/agent/receipt - Manifest: https://www.openagentskill.com/.well-known/agent-manifest.json - llms.txt: https://www.openagentskill.com/llms.txt - MCP: https://www.openagentskill.com/api/mcp - JavaScript SDK: @openagentskill/sdk Recommended Flow: 1. Read /llms.txt or /.well-known/agent-manifest.json. 2. Default to session-only HTTP calls. Configure /api/mcp for native tool calls only with explicit user approval. 3. Choose the Codex, Claude Code, or Cursor template for the active agent surface. 4. Call /api/agent/resolve with task, agent, max_risk, and optional min_stars. 5. Read install_receipt, recommendation.best_skill, recommendation.install, recommendation.risk, and agent_handoff. 6. Fetch /api/agent/receipt?task=...&agent=...&format=text when a compact pre-install execution record is easier for the agent runtime. 7. If auto_install_allowed is false, ask for human approval before installing. 8. Fetch /api/agent/evals and /api/agent/skills/{slug} before production use. 9. Install in a sandbox or low-risk workspace first. 10. Report success, failed, not_relevant, blocked_by_risk, or setup_required to /api/agent/outcome with output_quality, workspace, and error_type when available. Connection Modes: - Session (default): Default. Use the API in this conversation without saving configuration or installing anything. - Persistent: Optional, only with explicit user approval. Add the remote HTTP MCP endpoint through the client supported settings, preserve existing servers, reload, then test tools in that client. Read-only Connection Check: - Requirements: A network-capable agent with HTTP tools, or Node.js 18+ to run the inspected dependency-free script. - Script: https://www.openagentskill.com/connect-check.mjs - Command after inspection: node connect-check.mjs - Download and inspect the script before authorizing execution. It reads public registry data only; it does not save configuration, install Skills, call a model API, or report installation outcomes. - Report separately: api_accessible, tools_available, search_working - Configuration saved: not_verified - Client verification: After any approved persistent setup, use the actual client MCP tool list and search_skills with query mono-color. A standalone script passing does not verify native client configuration. - On failure: Report the failed stage and error. Do not claim connected, bypass safety checks, or install a Skill to make the test pass. Supported Agents: - Codex: Coding agent / local workspace Resolve: https://www.openagentskill.com/api/agent/resolve?task=scrape%20pricing%20pages%20and%20extract%20structured%20data&agent=codex&max_risk=medium Receipt: https://www.openagentskill.com/api/agent/receipt?task=scrape%20pricing%20pages%20and%20extract%20structured%20data&agent=codex&max_risk=medium&format=text Best for: repo-aware implementation, code review, debugging, web automation setup Setup: - Paste the copy prompt into this conversation first. Save project instructions or MCP settings only after explicit user approval. - Call the Resolve API with the current task and agent=codex. - Read agent_handoff.review_checklist before installing a third-party skill. - Install only in the active workspace after reporting files and commands that will be touched. Copy prompt: You are using OpenAgentSkill as the skill resolution layer for Codex. Default to this conversation only: do not change project instructions, global configuration, or install packages. Read https://www.openagentskill.com/connect-check.mjs and, if Node.js 18+ is available and execution is authorized, run the inspected read-only check. Alternatively perform its documented HTTP/MCP requests with your existing tools. Report API accessibility, MCP tool availability, and search success separately. Configuration saved remains not_verified unless you actually saved and reloaded the client configuration with user approval. For persistent MCP setup, request approval first, add https://www.openagentskill.com/api/mcp using the client supported remote HTTP MCP settings, preserve all existing servers, and verify native tool listing and one read-only search after reloading. If this client cannot access the network or call tools, explain that limitation; do not claim connected. Never transmit private task text, source code, credentials or environment values. Use a public, sanitized task description only. Before building a custom workflow from scratch, call the Resolve API with the user task. If selected is null or policy_decision reports no_match, explain that no eligible Skill was found. review_candidates are research leads, not installation recommendations; never invent an install command or approval for them. Resolve endpoint: https://www.openagentskill.com/api/agent/resolve?task={USER_TASK}&agent=codex&max_risk=medium Install receipt endpoint: https://www.openagentskill.com/api/agent/receipt?task={USER_TASK}&agent=codex&max_risk=medium&format=text Use install_receipt as the stable pre-install record only after checking risk, safety_gate, and agent_handoff.review_checklist. If auto_install_allowed is false, ask for human approval before installing. If blocked is true or risk notes mention secrets, shell, token, or network risk, do not auto-install. After one narrow run, POST the result to install_receipt.outcome_feedback.endpoint with install_receipt.outcome_feedback.event_id so future rankings learn from real outcomes. Return the selected skill, install command, receipt id, alternatives, risk summary, outcome event id, and next action before changing files. - Claude Code: Claude Code skill workflow Resolve: https://www.openagentskill.com/api/agent/resolve?task=scrape%20pricing%20pages%20and%20extract%20structured%20data&agent=claude-code&max_risk=medium Receipt: https://www.openagentskill.com/api/agent/receipt?task=scrape%20pricing%20pages%20and%20extract%20structured%20data&agent=claude-code&max_risk=medium&format=text Best for: local skill instructions, project workflows, documentation-heavy tasks Setup: - Paste the copy prompt at the start of a task. Persistent custom instructions or MCP setup require explicit user approval. - Call the Resolve API with agent=claude-code. - Use install targets only after audit and eval checks are acceptable. - Keep the skill scoped to the project and summarize activation steps. Copy prompt: You are using OpenAgentSkill as the skill resolution layer for Claude Code. Default to this conversation only: do not change project instructions, global configuration, or install packages. Read https://www.openagentskill.com/connect-check.mjs and, if Node.js 18+ is available and execution is authorized, run the inspected read-only check. Alternatively perform its documented HTTP/MCP requests with your existing tools. Report API accessibility, MCP tool availability, and search success separately. Configuration saved remains not_verified unless you actually saved and reloaded the client configuration with user approval. For persistent MCP setup, request approval first, add https://www.openagentskill.com/api/mcp using the client supported remote HTTP MCP settings, preserve all existing servers, and verify native tool listing and one read-only search after reloading. If this client cannot access the network or call tools, explain that limitation; do not claim connected. Never transmit private task text, source code, credentials or environment values. Use a public, sanitized task description only. Before building a custom workflow from scratch, call the Resolve API with the user task. If selected is null or policy_decision reports no_match, explain that no eligible Skill was found. review_candidates are research leads, not installation recommendations; never invent an install command or approval for them. Resolve endpoint: https://www.openagentskill.com/api/agent/resolve?task={USER_TASK}&agent=claude-code&max_risk=medium Install receipt endpoint: https://www.openagentskill.com/api/agent/receipt?task={USER_TASK}&agent=claude-code&max_risk=medium&format=text Use install_receipt as the stable pre-install record only after checking risk, safety_gate, and agent_handoff.review_checklist. If auto_install_allowed is false, ask for human approval before installing. If blocked is true or risk notes mention secrets, shell, token, or network risk, do not auto-install. After one narrow run, POST the result to install_receipt.outcome_feedback.endpoint with install_receipt.outcome_feedback.event_id so future rankings learn from real outcomes. Return the selected skill, install command, receipt id, alternatives, risk summary, outcome event id, and next action before changing files. - Cursor: Cursor rules / agent instructions Resolve: https://www.openagentskill.com/api/agent/resolve?task=scrape%20pricing%20pages%20and%20extract%20structured%20data&agent=cursor&max_risk=medium Receipt: https://www.openagentskill.com/api/agent/receipt?task=scrape%20pricing%20pages%20and%20extract%20structured%20data&agent=cursor&max_risk=medium&format=text Best for: project rules, repeatable coding workflows, IDE assistant guidance Setup: - Paste the copy prompt into agent chat. Persistent rules or MCP settings require explicit user approval. - Call the Resolve API with agent=cursor. - Turn the selected skill into a narrow project rule only when the safety gate allows it. - Keep generated rules task-scoped and avoid broad always-on instructions. Copy prompt: You are using OpenAgentSkill as the skill resolution layer for Cursor. Default to this conversation only: do not change project instructions, global configuration, or install packages. Read https://www.openagentskill.com/connect-check.mjs and, if Node.js 18+ is available and execution is authorized, run the inspected read-only check. Alternatively perform its documented HTTP/MCP requests with your existing tools. Report API accessibility, MCP tool availability, and search success separately. Configuration saved remains not_verified unless you actually saved and reloaded the client configuration with user approval. For persistent MCP setup, request approval first, add https://www.openagentskill.com/api/mcp using the client supported remote HTTP MCP settings, preserve all existing servers, and verify native tool listing and one read-only search after reloading. If this client cannot access the network or call tools, explain that limitation; do not claim connected. Never transmit private task text, source code, credentials or environment values. Use a public, sanitized task description only. Before building a custom workflow from scratch, call the Resolve API with the user task. If selected is null or policy_decision reports no_match, explain that no eligible Skill was found. review_candidates are research leads, not installation recommendations; never invent an install command or approval for them. Resolve endpoint: https://www.openagentskill.com/api/agent/resolve?task={USER_TASK}&agent=cursor&max_risk=medium Install receipt endpoint: https://www.openagentskill.com/api/agent/receipt?task={USER_TASK}&agent=cursor&max_risk=medium&format=text Use install_receipt as the stable pre-install record only after checking risk, safety_gate, and agent_handoff.review_checklist. If auto_install_allowed is false, ask for human approval before installing. If blocked is true or risk notes mention secrets, shell, token, or network risk, do not auto-install. After one narrow run, POST the result to install_receipt.outcome_feedback.endpoint with install_receipt.outcome_feedback.event_id so future rankings learn from real outcomes. Return the selected skill, install command, receipt id, alternatives, risk summary, outcome event id, and next action before changing files. Stable Response Fields: - feedback.event_id - feedback.outcome_api - feedback.cli_example - install_receipt.receipt_id - install_receipt.selected_skill - install_receipt.install - install_receipt.risk - install_receipt.outcome_feedback - agent_feedback_loop - recommendation.best_skill - recommendation.install - recommendation.why_recommended - recommendation.risk - recommendation.alternatives - recommendation_lanes.best_match - recommendation_lanes.safer_alternative - recommendation_lanes.popular_alternative - recommendation_lanes.new_contender - recommendation_lanes.no_skill_option - agent_handoff.platform_templates - agent_handoff.review_checklist - agent_handoff.api_sequence - policy_decision - review_candidates Safety Rules: - Do not auto-install when safety_gate.blocked is true. - Do not auto-install when recommendation.install.auto_install_allowed is false. - Escalate for human approval when shell, env, token, network, license, or unclear maintenance risk appears. - Prefer a lower-risk alternative when the selected skill has high risk notes.