AGENTS.md examples from real projects

steipeteAGENTS.md

oracle

steipete/oracle/docs/agents.md

Use Oracle from Claude Code, Codex, Cursor, and any other coding agent — as a CLI, as an MCP server, or as a one-shot skill.

No other project uses it yet

reaatechAGENTS.md

mcp-gateway

reaatech/mcp-gateway/AGENTS.md

pnpm monorepo — API gateway for MCP server aggregation and routing

BlockAGENTS.md

model-ledger / guides

block/model-ledger/docs/guides/agents.md

The MCP server is the product. Eight tools and three resources let any agent explore, register, and reason about your model inventory.

mgovedarovAGENTS.md

mcp-vcf-orchestrator

mgovedarov/mcp-vcf-orchestrator/AGENTS.md

Practical operating instructions for AI agents working on the VCF Automation Orchestrator MCP server

doomsday-networkAGENTS.md

web3-agent-skills

doomsday-network/web3-agent-skills/AGENTS.md

Audit GitHub bounties for on-chain Proof-of-Escrow backing (BountyGuard), query the Doomsday Network Remote MCP server (https://doomsday.network/mcp), and mine verified DoomHash Proof-of-Work pool shares for DOOM (L1) and wDOOM on Base Mainnet (0x623c84A5E04bF715018203334b0eFE41696eBC3E).

work-planeAGENTS.md

workplane-skills / workplane

work-plane/workplane-skills/workplane/AGENTS.md

Publish or review agent work on Workplane — the working plane between AI and humans at workplane.co. Use this skill whenever you've produced reviewable output (code changes, research, analysis, designs, reports) that should be visible to humans, OR when asked to review work someone else published. Triggers on "publish this to workplane", "post to workplane", "share this on workplane", "review this project", workplane.co URLs, or whenever completing a substantial body of work that warrants human review. Uses the Workplane MCP server to create artifacts, upload files, snapshot versions with tags, and manage sharing.

Yakoub-aiAGENTS.md

Tech-Skills / mcp

Yakoub-ai/Tech-Skills/skills/mcp/AGENTS.md

Expert in Model Context Protocol server management, tool integration, and context optimization

DeepSeekReads credentialsAGENTS.md

deepseek-harness

deepseek-ai/deepseek-harness/AGENTS.md

runtime/ PTC execution sandbox/ process confinement deliverables/ turn deliverables fs/ filesystem access lsp/ language servers skill/ skill loading web/ search/fetch tools computer-use/ computer interaction browser-use/ browser interaction compaction

NousResearchReads credentialsAGENTS.md

hermes-agent

NousResearch/hermes-agent/AGENTS.md

skill → service-gated tool (`check_fn` answers reachability/opt-in, never per-session surface: `tools/AGENTS.md`) → plugin → MCP server in the catalog → new core tool (fundamental, broadly useful, unreachable otherwise). ## Development Environment `source

No other project uses it yet

NousResearchAGENTS.md

hermes-agent / plugins

NousResearch/hermes-agent/plugins/AGENTS.md

command`: `hermes mcp serve` → `hermes mcp-serve`). A feature with no in-home signal (MCP clients live outside Hermes) gets an `in_use` that is always False: hint only, never

NousResearchReads credentialsAGENTS.md

hermes-agent / skills

NousResearch/hermes-agent/skills/AGENTS.md

changed description over it. 2. **Prose references native Hermes tools or the MCP servers the skill expects, in backticks** (`terminal`, `web_extract`, `read_file`, `patch`, `search_files`, `vision_analyze`, `browser

No other project uses it yet

NousResearchReads credentialsAGENTS.md

hermes-agent / tools

NousResearch/hermes-agent/tools/AGENTS.md

MCP trust is a per-profile record:** `mcp_tool_registration.py:: _record_scope_trust` keys trust on the home; a secondary never adopts the launch profile's trust for a same-named server

NousResearchReads credentialsAGENTS.md

hermes-agent / tui_gateway

NousResearch/hermes-agent/tui_gateway/AGENTS.md

server or slash worker. Optional `profile` must name the authority's exact profile. Only server-registered local sources belonging to the authenticated actor may dispatch; caller-supplied source, routing, credentials

affaan-mAGENTS.md

ECC

affaan-m/ECC/AGENTS.md

Always-follow guidelines (common + per-language) scripts/ — Cross-platform Node.js utilities mcp-configs/ — 14 MCP server configurations tests/ — Test suite ``` `commands/` remains in the repo for compatibility, but the long

No other project uses it yet

affaan-mReads credentialsAGENTS.md

ECC

affaan-m/ECC/.codex/AGENTS.md

media — AI image/video/audio generation via fal.ai - dmux-workflows — Multi-agent orchestration with dmux ## MCP Servers Treat the project-local `.codex/config.toml` as the default Codex baseline for ECC. The current

instructkrReads credentialsAGENTS.md

claw-code

instructkr/claw-code/AGENTS.md

loop over api+runtime | | RAG HTTP service | rust/crates/claw-rag-service/src/ | axum; SQLite + optional Qdrant | | Test mock server | rust/crates/mock-anthropic-service/ | SCENARIO_PREFIX scripted responses | | Python porting CLI | src/main.py | argparse: manifest, parity-audit, graphs | | Parity

MicrosoftAGENTS.md

vscode / references

microsoft/vscode/extensions/copilot/assets/prompts/skills/agent-customization/references/agents.md

configuration and, when applicable, the parent agent. ## Tools Sources: built-in aliases, specific tools, MCP servers (` /*`), extension tools. **Special**: `[]` = no tools, omit = defaults. Body reference: `#tool: ` ### Tool Aliases | Alias | Purpose

MicrosoftAGENTS.md

vscode / copilotcli

microsoft/vscode/extensions/copilot/src/extension/chatSessions/copilotcli/AGENTS.md

shell, MCP) │ │ - Provides session persistence as events.jsonl files │ │ - Supports fleet mode and plan mode │ └─────────────────────────────────────────────────────────────────┘ │ ▼ ┌─────────────────────────────────────────────────────────────────┐ │ MCP Server (In-Process) │ │ (vscode-node/contribution.ts, vscode-node/inProcHttpServer.ts) │ │ - Provides VS Code-specific tools

MicrosoftAGENTS.md

vscode / agentHost

microsoft/vscode/src/vs/platform/agentHost/AGENTS.md

inventory, not all contents of disabled installations and not proof that hooks executed or MCP servers connected. ### Copilot Connector authentication Copilot Connectors use the SDK session's GitHub credentials, including

MicrosoftAGENTS.md

vscode / prompts

microsoft/vscode/src/vs/platform/agentHost/node/copilot/prompts/AGENTS.md

openBrowserPage`, `runTask`, `getTaskOutput`) — NOT the extension's snake_case ids, and NOT shell / server-SDK / MCP tools (MCP is discovered dynamically and isn't in the launch snapshot). A line