[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M6DBKW56HYZZXZSJSJWT12C1":3,"skill-same:01M6DBKW56HYZZXZSJSJWT12C1":105},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":16,"categories":17,"keywords":23,"publishedAt":44,"linkOnly":45,"markdownContent":46,"markdownBody":47,"frontMatter":48,"refName":49,"commitSha":50,"skillDirectory":51,"forksCount":14,"repositoryHtmlUrl":52,"repositoryHomepageUrl":6,"ownerHtmlUrl":53,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":54,"install":55,"resources":68},"01M6DBKW56HYZZXZSJSJWT12C1","multi-ai-mingle",null,"Use when the user wants seamless local collaboration between Claude Code and OpenAI Codex on the same machine - fanning out parallel Codex coding threads from Claude as native MCP tool calls via the official codex mcp-server, and running a live symmetric two-way chat and file-lease channel between the two agents through mcp_agent_mail. Covers mounting codex-as-MCP at user scope, standing up the shared agent-mail http channel, verifying it before use, and the sequenced adoption waves. Triggers include mingle Claude and Codex, run Codex from Claude in parallel, set up the agent mail channel, replace AgentBridge, or coordinate the two local AI agents. Not for single-agent work or remote or cloud multi-agent orchestration.","Prekzursil\u002Fagent-skills-toolchain","Prekzursil","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F54636077?v=4","01M6DACZ1PCQHD99SCR1YXG28B","Andrei Mihai Visalon","01M6DACZ670WZF535V3WYYSS4Z",0,"MIT","https:\u002F\u002Fgithub.com\u002FPrekzursil\u002Fagent-skills-toolchain\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Fmulti-ai-mingle\u002FSKILL.md",[18],{"categoryNo":19,"categoryName":20,"taxonomyType":21,"taxonomyTypeDesc":22},"01JBSEEDCAT000000000000023","Skill与智能体开发","SCENARIO","使用场景",[24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43],"multi ai mingle","up","openai codex","live","shared","local","mail","coordinate","local ai","agent orchestration","MCP","work","SET","server","code","run","official","local ai agents","orchestration","ai agents","2026-11-10 10:12:30",false,"---\nname: multi-ai-mingle\ndescription: Use when the user wants seamless local collaboration between Claude Code and OpenAI Codex on the same machine - fanning out parallel Codex coding threads from Claude as native MCP tool calls via the official codex mcp-server, and running a live symmetric two-way chat and file-lease channel between the two agents through mcp_agent_mail. Covers mounting codex-as-MCP at user scope, standing up the shared agent-mail http channel, verifying it before use, and the sequenced adoption waves. Triggers include mingle Claude and Codex, run Codex from Claude in parallel, set up the agent mail channel, replace AgentBridge, or coordinate the two local AI agents. Not for single-agent work or remote or cloud multi-agent orchestration.\n---\n\n# multi-ai-mingle\n\nSeamless LOCAL mingling of Claude Code and OpenAI Codex on one Windows 11 machine. Two\ncomplementary rails:\n\n- **Fan-out rail (Claude drives Codex):** the official `codex mcp-server` (stdio) mounted into\n  Claude at user scope. Claude subagents call `mcp__codex__codex` \u002F `mcp__codex__codex_reply` as\n  native tools, so one Claude orchestrator can run several parallel Codex coding threads.\n- **Live channel rail (peer chat + coordination):** `mcp_agent_mail` (2k-star, maintained;\n  FastMCP + Git + SQLite) as a shared http server both agents connect to — symmetric inboxes,\n  threads, and advisory file leases. This replaces the Windows-broken AgentBridge.\n\nKeep the two rails distinct: fan-out is command-and-control (Claude tasks Codex); the channel is\npeer messaging + file-lease coordination that either side reads and writes.\n\n## When to use\n- The user wants Claude and Codex to collaborate locally, in parallel, with a real back-channel.\n- The user asks to \"run Codex from Claude\", \"set up the agent-mail channel\", or \"replace AgentBridge\".\n\n## When NOT to use\n- Single-agent tasks (just use Claude, or Codex, directly).\n- Remote \u002F cloud multi-agent fleets (different tooling).\n\n---\n\n## Real mcp_agent_mail API sequence (VERIFIED 2026-07-11 — the prep draft under-specified this)\nThe channel is not \"just send_message\". The working order, proven by the verify probe on Windows:\n1. `ensure_project(human_key=\u003Cabs path>)` FIRST — creates the namespace. (The **other** tools take\n   that same path as `project_key`; only ensure_project uses `human_key`.)\n2. `register_agent(project_key, program=\u003Crole>, model)` per agent. **Bare role names `claude`\u002F`codex`\n   are rejected** — the server assigns an adjective+noun codename (e.g. `BlackCompass`) AND returns a\n   `registration_token`. Capture BOTH from the result.\n3. First contact is **gated (anti-spam)**: `request_contact(from_agent, to_agent, registration_token)`\n   then the recipient `respond_contact(to_agent, from_agent, accept=true, registration_token)`.\n4. Then `send_message(sender_name, sender_token, to=[...])`, `fetch_inbox(agent_name, registration_token)`,\n   `file_reservation_paths(agent_name, registration_token, paths)` — each call carries the agent's token\n   (a single MCP session can only be auto-bound to one agent; pass tokens explicitly for the rest).\nIn real use each of Claude\u002FCodex has its OWN session bound to its OWN agent, so it authenticates once and\nomits the token thereafter; the two-in-one-session token-passing above is a probe artifact.\n\n## Wave A — mount + channel (STATUS: DONE + VERIFIED 2026-07-11)\n- ✅ codex mcp-server mounted (user scope, Connected). ✅ mcp_agent_mail installed+pinned+probe-GREEN\n  (symmetric both ways + file-lease). ✅ AgentBridge retired (`claude plugin disable agentbridge@agentbridge`).\n- ⚠️ Retiring AgentBridge also removed its `get_budget` tool → the CLAUDE.md budget-pacing protocol is\n  now inert (was already dormant — daemon down). Re-enable the plugin if budget-pacing is wanted back.\n\n\n1. **Mount codex-as-MCP (user scope).** See `MOUNT.md`.\n   - `claude mcp add --scope user codex -- codex mcp-server`\n   - Set a long tool timeout (Codex turns take minutes): `MCP_TOOL_TIMEOUT=600000` and\n     `MCP_TIMEOUT=30000` in `~\u002F.claude\u002Fsettings.json` env, or bake a per-server `timeout` via\n     `claude mcp add-json`.\n   - Verify: `claude mcp get codex`; ask Claude to run a trivial codex turn.\n\n2. **Stand up the shared agent-mail http channel.** See `MCP_AGENT_MAIL.md`.\n   - Install (native Windows): clone, `uv python install 3.14`, `uv venv -p 3.14`, `uv sync`.\n   - Run ONE shared server on the canonical port **8812**: `uv run python -m mcp_agent_mail.cli\n     serve-http --host 127.0.0.1 --port 8812` (health: `\u002Fapi\u002Fhealth` returns `{\"status\":\"ok\"}`).\n     **Do NOT use 8765 — that is the LIVE basic-memory port; never collide.** On this box the\n     server is kept always-on by the `MCPAgentMailSupervisor` logon task; force a restart with\n     `Start-ScheduledTask -TaskName MCPAgentMailSupervisor`.\n   - Register the SAME url in both agents: Claude via `claude mcp add --transport http --scope user\n     agent-mail http:\u002F\u002F127.0.0.1:8812\u002Fmcp --header \"Authorization: Bearer TOKEN\"`; Codex via a\n     `[mcp_servers.agent-mail]` http block in `~\u002F.codex\u002Fconfig.toml` with `AGENT_NAME` under\n     `env_vars`. Identity: `claude` on Claude's side, `codex` on Codex's side.\n   - Tools: `send_message`, `fetch_inbox`, `reply_message`, `file_reservation_paths`,\n     `release_file_reservations` (plus `register_agent`, `whois`, `list_agents`, `search_messages`).\n\n3. **VERIFY BEFORE RETIRING AGENTBRIDGE.** Run `VERIFY_PLAN.ps1` (native Windows; POSIX mirror\n   `VERIFY_PLAN.sh`). It starts an isolated server, proves symmetric claude to codex and codex to\n   claude message visibility, and a file-lease acquire\u002Frelease round-trip, printing SUCCESS: \u002F\n   FAILED: lines. Only after all-green: commit the configs and retire AgentBridge.\n\n## Wave B — teammate shim (LATER)\nFork the `claude-anyteam` `CLAUDE_CODE_TEAMMATE_COMMAND` shim into `~\u002F.agents\u002Fharness` so Codex can\nbe launched as a first-class Claude Code teammate (TeamCreate \u002F SendMessage surface), not only an\nMCP tool. Gives dispatch parity with native subagents while the agent-mail channel carries the\ncoordination traffic.\n\n## Wave C — GUI tab (LATER)\nAdd a GUI tab (agent-mail ships a server-rendered web UI + a TUI console) so the human can watch and\ninject into the Claude\u002FCodex thread and see live file leases.\n\n## Wave D — upstream contribution (LATER)\nContribute the AgentBridge fixes upstream (issues #231 \u002F #223) so the learnings from replacing it are\nreturned to that project rather than just forked away.\n\n---\n\n## Files in this skill\n- `MOUNT.md` — exact codex-as-MCP mount + timeout + verify + reverse.\n- `MCP_AGENT_MAIL.md` — verified repo, install, dual-agent registration, tool surface, Windows caveats.\n- `VERIFY_PLAN.ps1` \u002F `VERIFY_PLAN.sh` \u002F `agent_mail_probe.py` — self-contained symmetric + file-lease smoke test.\n\n## Guardrails\n- One shared `serve-http`, not per-client stdio, or the two agents will not see each other's mail.\n- Bind agent-mail to `127.0.0.1` and keep the bearer token; do not expose the port.\n- **Port discipline:** agent-mail is canonically **8812** (`\u002Fapi\u002Fhealth` -> `{\"status\":\"ok\"}`);\n  **8765 belongs to the LIVE basic-memory MCP server** on this box — never move agent-mail onto\n  it or the two servers collide.\n- Codex turns are long — keep `MCP_TOOL_TIMEOUT` generous or tool calls get cut off mid-turn.\n- Python 3.14 + native-Windows wheel builds are the residual risk; the verify step is the gate.\n- **Agent identity:** bare role names (`claude`\u002F`codex`) are REJECTED in strict mode — register with\n  the role in `program` and address agents by the server-ASSIGNED adjective+noun codename (or pick an\n  explicit VALID name like `BlueLake`). Call `ensure_project(human_key=\u003Cabs path>)` BEFORE any\n  register\u002Fsend\u002Flease (the rest of the tools take that same path as `project_key`).\n\n## License + adoption posture (owner directive 2026-07-11)\n- **License:** mcp_agent_mail is **MIT + an OpenAI\u002FAnthropic rider** — it grants no rights to those two\n  vendors\u002Faffiliates and its \"use\" definition forbids incorporating the tool into any **ML training or\n  eval pipeline**. For this owner's **private, local** two-agent coordination that is fine (not a\n  restricted party, not an eval harness). **Never** fold mcp_agent_mail into an eval\u002Ftraining harness.\n- **Pinned + sandboxed + swappable** (the adopt-but-keep-options-open call): vendored at\n  `~\u002F.local\u002Fopt\u002Fmcp_agent_mail`, **pinned to commit `35e774fa9ae636c6e1662ab7925d7e68938bb718`**\n  (post-fix for the Windows lock-wedge #91), run **offline** (127.0.0.1, no JWT\u002Fshare\u002Fcloud features).\n  The live-channel layer is a thin swappable rail — it **converges onto the Claude Code Agent Teams\n  mailbox** once the Wave-B teammate shim lands, so there is no lock-in.\n- **Verified facts (this machine, 2026-07-11):** `codex mcp-server` mounted at user scope and reports\n  **Connected**; MCP timeouts already generous (`MCP_TOOL_TIMEOUT=300000`, `MCP_TIMEOUT=120000`) — no\n  change needed. Note the verify probe uses one connection driving both identities (proves the shared\n  store routes + leases); a two-process concurrent test is the deeper follow-up (#91's fix makes it low-risk).\n","# multi-ai-mingle\n\nSeamless LOCAL mingling of Claude Code and OpenAI Codex on one Windows 11 machine. Two\ncomplementary rails:\n\n- **Fan-out rail (Claude drives Codex):** the official `codex mcp-server` (stdio) mounted into\n  Claude at user scope. Claude subagents call `mcp__codex__codex` \u002F `mcp__codex__codex_reply` as\n  native tools, so one Claude orchestrator can run several parallel Codex coding threads.\n- **Live channel rail (peer chat + coordination):** `mcp_agent_mail` (2k-star, maintained;\n  FastMCP + Git + SQLite) as a shared http server both agents connect to — symmetric inboxes,\n  threads, and advisory file leases. This replaces the Windows-broken AgentBridge.\n\nKeep the two rails distinct: fan-out is command-and-control (Claude tasks Codex); the channel is\npeer messaging + file-lease coordination that either side reads and writes.\n\n## When to use\n- The user wants Claude and Codex to collaborate locally, in parallel, with a real back-channel.\n- The user asks to \"run Codex from Claude\", \"set up the agent-mail channel\", or \"replace AgentBridge\".\n\n## When NOT to use\n- Single-agent tasks (just use Claude, or Codex, directly).\n- Remote \u002F cloud multi-agent fleets (different tooling).\n\n---\n\n## Real mcp_agent_mail API sequence (VERIFIED 2026-07-11 — the prep draft under-specified this)\nThe channel is not \"just send_message\". The working order, proven by the verify probe on Windows:\n1. `ensure_project(human_key=\u003Cabs path>)` FIRST — creates the namespace. (The **other** tools take\n   that same path as `project_key`; only ensure_project uses `human_key`.)\n2. `register_agent(project_key, program=\u003Crole>, model)` per agent. **Bare role names `claude`\u002F`codex`\n   are rejected** — the server assigns an adjective+noun codename (e.g. `BlackCompass`) AND returns a\n   `registration_token`. Capture BOTH from the result.\n3. First contact is **gated (anti-spam)**: `request_contact(from_agent, to_agent, registration_token)`\n   then the recipient `respond_contact(to_agent, from_agent, accept=true, registration_token)`.\n4. Then `send_message(sender_name, sender_token, to=[...])`, `fetch_inbox(agent_name, registration_token)`,\n   `file_reservation_paths(agent_name, registration_token, paths)` — each call carries the agent's token\n   (a single MCP session can only be auto-bound to one agent; pass tokens explicitly for the rest).\nIn real use each of Claude\u002FCodex has its OWN session bound to its OWN agent, so it authenticates once and\nomits the token thereafter; the two-in-one-session token-passing above is a probe artifact.\n\n## Wave A — mount + channel (STATUS: DONE + VERIFIED 2026-07-11)\n- ✅ codex mcp-server mounted (user scope, Connected). ✅ mcp_agent_mail installed+pinned+probe-GREEN\n  (symmetric both ways + file-lease). ✅ AgentBridge retired (`claude plugin disable agentbridge@agentbridge`).\n- ⚠️ Retiring AgentBridge also removed its `get_budget` tool → the CLAUDE.md budget-pacing protocol is\n  now inert (was already dormant — daemon down). Re-enable the plugin if budget-pacing is wanted back.\n\n\n1. **Mount codex-as-MCP (user scope).** See `MOUNT.md`.\n   - `claude mcp add --scope user codex -- codex mcp-server`\n   - Set a long tool timeout (Codex turns take minutes): `MCP_TOOL_TIMEOUT=600000` and\n     `MCP_TIMEOUT=30000` in `~\u002F.claude\u002Fsettings.json` env, or bake a per-server `timeout` via\n     `claude mcp add-json`.\n   - Verify: `claude mcp get codex`; ask Claude to run a trivial codex turn.\n\n2. **Stand up the shared agent-mail http channel.** See `MCP_AGENT_MAIL.md`.\n   - Install (native Windows): clone, `uv python install 3.14`, `uv venv -p 3.14`, `uv sync`.\n   - Run ONE shared server on the canonical port **8812**: `uv run python -m mcp_agent_mail.cli\n     serve-http --host 127.0.0.1 --port 8812` (health: `\u002Fapi\u002Fhealth` returns `{\"status\":\"ok\"}`).\n     **Do NOT use 8765 — that is the LIVE basic-memory port; never collide.** On this box the\n     server is kept always-on by the `MCPAgentMailSupervisor` logon task; force a restart with\n     `Start-ScheduledTask -TaskName MCPAgentMailSupervisor`.\n   - Register the SAME url in both agents: Claude via `claude mcp add --transport http --scope user\n     agent-mail http:\u002F\u002F127.0.0.1:8812\u002Fmcp --header \"Authorization: Bearer TOKEN\"`; Codex via a\n     `[mcp_servers.agent-mail]` http block in `~\u002F.codex\u002Fconfig.toml` with `AGENT_NAME` under\n     `env_vars`. Identity: `claude` on Claude's side, `codex` on Codex's side.\n   - Tools: `send_message`, `fetch_inbox`, `reply_message`, `file_reservation_paths`,\n     `release_file_reservations` (plus `register_agent`, `whois`, `list_agents`, `search_messages`).\n\n3. **VERIFY BEFORE RETIRING AGENTBRIDGE.** Run `VERIFY_PLAN.ps1` (native Windows; POSIX mirror\n   `VERIFY_PLAN.sh`). It starts an isolated server, proves symmetric claude to codex and codex to\n   claude message visibility, and a file-lease acquire\u002Frelease round-trip, printing SUCCESS: \u002F\n   FAILED: lines. Only after all-green: commit the configs and retire AgentBridge.\n\n## Wave B — teammate shim (LATER)\nFork the `claude-anyteam` `CLAUDE_CODE_TEAMMATE_COMMAND` shim into `~\u002F.agents\u002Fharness` so Codex can\nbe launched as a first-class Claude Code teammate (TeamCreate \u002F SendMessage surface), not only an\nMCP tool. Gives dispatch parity with native subagents while the agent-mail channel carries the\ncoordination traffic.\n\n## Wave C — GUI tab (LATER)\nAdd a GUI tab (agent-mail ships a server-rendered web UI + a TUI console) so the human can watch and\ninject into the Claude\u002FCodex thread and see live file leases.\n\n## Wave D — upstream contribution (LATER)\nContribute the AgentBridge fixes upstream (issues #231 \u002F #223) so the learnings from replacing it are\nreturned to that project rather than just forked away.\n\n---\n\n## Files in this skill\n- `MOUNT.md` — exact codex-as-MCP mount + timeout + verify + reverse.\n- `MCP_AGENT_MAIL.md` — verified repo, install, dual-agent registration, tool surface, Windows caveats.\n- `VERIFY_PLAN.ps1` \u002F `VERIFY_PLAN.sh` \u002F `agent_mail_probe.py` — self-contained symmetric + file-lease smoke test.\n\n## Guardrails\n- One shared `serve-http`, not per-client stdio, or the two agents will not see each other's mail.\n- Bind agent-mail to `127.0.0.1` and keep the bearer token; do not expose the port.\n- **Port discipline:** agent-mail is canonically **8812** (`\u002Fapi\u002Fhealth` -> `{\"status\":\"ok\"}`);\n  **8765 belongs to the LIVE basic-memory MCP server** on this box — never move agent-mail onto\n  it or the two servers collide.\n- Codex turns are long — keep `MCP_TOOL_TIMEOUT` generous or tool calls get cut off mid-turn.\n- Python 3.14 + native-Windows wheel builds are the residual risk; the verify step is the gate.\n- **Agent identity:** bare role names (`claude`\u002F`codex`) are REJECTED in strict mode — register with\n  the role in `program` and address agents by the server-ASSIGNED adjective+noun codename (or pick an\n  explicit VALID name like `BlueLake`). Call `ensure_project(human_key=\u003Cabs path>)` BEFORE any\n  register\u002Fsend\u002Flease (the rest of the tools take that same path as `project_key`).\n\n## License + adoption posture (owner directive 2026-07-11)\n- **License:** mcp_agent_mail is **MIT + an OpenAI\u002FAnthropic rider** — it grants no rights to those two\n  vendors\u002Faffiliates and its \"use\" definition forbids incorporating the tool into any **ML training or\n  eval pipeline**. For this owner's **private, local** two-agent coordination that is fine (not a\n  restricted party, not an eval harness). **Never** fold mcp_agent_mail into an eval\u002Ftraining harness.\n- **Pinned + sandboxed + swappable** (the adopt-but-keep-options-open call): vendored at\n  `~\u002F.local\u002Fopt\u002Fmcp_agent_mail`, **pinned to commit `35e774fa9ae636c6e1662ab7925d7e68938bb718`**\n  (post-fix for the Windows lock-wedge #91), run **offline** (127.0.0.1, no JWT\u002Fshare\u002Fcloud features).\n  The live-channel layer is a thin swappable rail — it **converges onto the Claude Code Agent Teams\n  mailbox** once the Wave-B teammate shim lands, so there is no lock-in.\n- **Verified facts (this machine, 2026-07-11):** `codex mcp-server` mounted at user scope and reports\n  **Connected**; MCP timeouts already generous (`MCP_TOOL_TIMEOUT=300000`, `MCP_TIMEOUT=120000`) — no\n  change needed. Note the verify probe uses one connection driving both identities (proves the shared\n  store routes + leases); a two-process concurrent test is the deeper follow-up (#91's fix makes it low-risk).\n",{"name":5,"description":7},"main","a298c929b6cbd662e1fce54302ce520d5ceea1a5",".agents\u002Fskills\u002Fmulti-ai-mingle","https:\u002F\u002Fgithub.com\u002FPrekzursil\u002Fagent-skills-toolchain","https:\u002F\u002Fgithub.com\u002FPrekzursil","2026-11-02 14:30:57",{"source":8,"skillName":5,"commands":56,"prompt":66,"zipAvailable":67},[57,60,63],{"tool":58,"command":59},"npx","npx skills add Prekzursil\u002Fagent-skills-toolchain --skill multi-ai-mingle",{"tool":61,"command":62},"bunx","bunx skills add Prekzursil\u002Fagent-skills-toolchain --skill multi-ai-mingle",{"tool":64,"command":65},"pnpm","pnpm dlx skills add Prekzursil\u002Fagent-skills-toolchain --skill multi-ai-mingle","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FPrekzursil\u002Fagent-skills-toolchain\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Fmulti-ai-mingle\u002FSKILL.md",true,[69,75,81,85,89,93,97,101],{"path":70,"resourceType":71,"resourceTypeDesc":72,"sizeBytes":73,"sourceHtmlUrl":74},"agent_mail_probe.py","SCRIPT","脚本",9460,"https:\u002F\u002Fgithub.com\u002FPrekzursil\u002Fagent-skills-toolchain\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Fmulti-ai-mingle\u002Fagent_mail_probe.py",{"path":76,"resourceType":77,"resourceTypeDesc":78,"sizeBytes":79,"sourceHtmlUrl":80},"CODEX_FANOUT.md","REFERENCE","参考文档",3653,"https:\u002F\u002Fgithub.com\u002FPrekzursil\u002Fagent-skills-toolchain\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Fmulti-ai-mingle\u002FCODEX_FANOUT.md",{"path":82,"resourceType":77,"resourceTypeDesc":78,"sizeBytes":83,"sourceHtmlUrl":84},"MCP_AGENT_MAIL.md",9488,"https:\u002F\u002Fgithub.com\u002FPrekzursil\u002Fagent-skills-toolchain\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Fmulti-ai-mingle\u002FMCP_AGENT_MAIL.md",{"path":86,"resourceType":77,"resourceTypeDesc":78,"sizeBytes":87,"sourceHtmlUrl":88},"MOUNT.md",4945,"https:\u002F\u002Fgithub.com\u002FPrekzursil\u002Fagent-skills-toolchain\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Fmulti-ai-mingle\u002FMOUNT.md",{"path":90,"resourceType":77,"resourceTypeDesc":78,"sizeBytes":91,"sourceHtmlUrl":92},"SKILL.alt.md",8639,"https:\u002F\u002Fgithub.com\u002FPrekzursil\u002Fagent-skills-toolchain\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Fmulti-ai-mingle\u002FSKILL.alt.md",{"path":94,"resourceType":77,"resourceTypeDesc":78,"sizeBytes":95,"sourceHtmlUrl":96},"SKILL.dot.alt.md",8719,"https:\u002F\u002Fgithub.com\u002FPrekzursil\u002Fagent-skills-toolchain\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Fmulti-ai-mingle\u002FSKILL.dot.alt.md",{"path":98,"resourceType":71,"resourceTypeDesc":72,"sizeBytes":99,"sourceHtmlUrl":100},"VERIFY_PLAN.ps1",4652,"https:\u002F\u002Fgithub.com\u002FPrekzursil\u002Fagent-skills-toolchain\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Fmulti-ai-mingle\u002FVERIFY_PLAN.ps1",{"path":102,"resourceType":71,"resourceTypeDesc":72,"sizeBytes":103,"sourceHtmlUrl":104},"VERIFY_PLAN.sh",3524,"https:\u002F\u002Fgithub.com\u002FPrekzursil\u002Fagent-skills-toolchain\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Fmulti-ai-mingle\u002FVERIFY_PLAN.sh",[106,128,153,180,194,209],{"skillNo":107,"name":108,"nameZh":6,"description":109,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":110,"categories":111,"keywords":112,"publishedAt":44},"01M6DF7D7478R55P46REYH4QAC","clarify-requirements","Resolve genuine pre-execution ambiguity by scanning repo context, generating 1-6 grounded options, and asking 1-3 focused questions via AskUserQuestion. NOT for mid-task use (injection risk); NOT when the goal is already unambiguous; use to disambiguate vague prompts before any file edits.","https:\u002F\u002Fgithub.com\u002FPrekzursil\u002Fagent-skills-toolchain\u002Fblob\u002Fmain\u002Ftools\u002Fworkspace-hooks\u002Fskills\u002Fclarify-requirements\u002FSKILL.md",[],[113,114,115,116,117,118,119,120,121,122,123,124,125,126,127],"clarify requirements","clarify","requirements","risk","options","task","prompts","goal","scanning","use","context","repo","injection","execution","file",{"skillNo":129,"name":130,"nameZh":6,"description":131,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":132,"categories":133,"keywords":135,"publishedAt":44},"01M6DF79CA4BHJS66J8EZDBPWJ","openmm-md","Run FREE, open-source, Windows-native GPU molecular dynamics via the openmm-md MCP (OpenMM + CUDA on the local NVIDIA GPU). Use for protein \u002F protein-ligand MD when you want a free stack or are on Windows without a Schrodinger license. Covers the full workflow (prepare, solvate, minimize, equilibrate, production, analyze, mmgbsa) AND an honest decision guide vs Schrodinger Desmond, including how to run Desmond legitimately on Linux or a cloud GPU with a real license. This is a deliberately LESSER free alternative and does NOT replace the schrodinger MCP tools. Do NOT use it to pick a free tool when a validated FEP+, WaterMap, IFD-MD, or Prime MM-GBSA result is actually required.","https:\u002F\u002Fgithub.com\u002FPrekzursil\u002Fagent-skills-toolchain\u002Fblob\u002Fmain\u002Ftools\u002Fmcp\u002Fopenmm-md-mcp\u002Fskill\u002Fopenmm-md\u002FSKILL.md",[134],{"categoryNo":19,"categoryName":20,"taxonomyType":21,"taxonomyTypeDesc":22},[136,137,138,139,140,141,142,143,144,145,146,147,148,149,122,150,151,29,152,34],"openmm md","how","decision","pick","linux","analyze","tool","stack","open","NVIDIA","gpu","license","workflow","windows","free","molecular dynamics","do",{"skillNo":154,"name":155,"nameZh":6,"description":156,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":157,"categories":158,"keywords":165,"publishedAt":44},"01M6DF76C1MG85VEWP8VZCBNMN","visual-review","Take screenshots of web pages and UI using Playwright for visual review and iteration","https:\u002F\u002Fgithub.com\u002FPrekzursil\u002Fagent-skills-toolchain\u002Fblob\u002Fmain\u002Ftools\u002Feditor\u002Fmetaswarm\u002Fskills\u002Fvisual-review\u002FSKILL.md",[159,162],{"categoryNo":160,"categoryName":161,"taxonomyType":21,"taxonomyTypeDesc":22},"01JBSEEDCAT000000000000003","演示文稿",{"categoryNo":163,"categoryName":164,"taxonomyType":21,"taxonomyTypeDesc":22},"01JBSEEDCAT000000000000008","测试与质量",[166,167,168,169,170,171,172,173,174,175,137,176,177,178,179],"review","visual","visual review","pages","web","false","iteration","take","check","slides","ui","screenshot","playwright","screenshots",{"skillNo":181,"name":182,"nameZh":6,"description":183,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":184,"categories":185,"keywords":186,"publishedAt":44},"01M6DF73AQVCBZKEN0YXEX9FXP","status","Diagnostic status report — shows metaswarm installation state, project setup, and potential issues","https:\u002F\u002Fgithub.com\u002FPrekzursil\u002Fagent-skills-toolchain\u002Fblob\u002Fmain\u002Ftools\u002Feditor\u002Fmetaswarm\u002Fskills\u002Fstatus\u002FSKILL.md",[],[182,187,188,189,190,191,192,193],"setup","report","project","installation","project setup","issues","state",{"skillNo":195,"name":196,"nameZh":6,"description":197,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":198,"categories":199,"keywords":201,"publishedAt":44},"01M6DF6X6C2XB0P6GX76P7TTYV","start","Use when starting work on any task, when the user mentions metaswarm, or when the user wants to begin tracked development work","https:\u002F\u002Fgithub.com\u002FPrekzursil\u002Fagent-skills-toolchain\u002Fblob\u002Fmain\u002Ftools\u002Feditor\u002Fmetaswarm\u002Fskills\u002Fstart\u002FSKILL.md",[200],{"categoryNo":19,"categoryName":20,"taxonomyType":21,"taxonomyTypeDesc":22},[196,202,118,203,204,205,122,206,35,207,208],"start issue","issue","ready","work on issue","user","work on","development",{"skillNo":210,"name":187,"nameZh":6,"description":211,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":212,"categories":213,"keywords":214,"publishedAt":44},"01M6DF6QMJ9AJKFKHHKRT005PQ","Interactive project setup — detects your project, configures metaswarm, writes project-local files","https:\u002F\u002Fgithub.com\u002FPrekzursil\u002Fagent-skills-toolchain\u002Fblob\u002Fmain\u002Ftools\u002Feditor\u002Fmetaswarm\u002Fskills\u002Fsetup\u002FSKILL.md",[],[187,189,29,215,191,216],"files","interactive"]