Add Agent Coordinator to the tools catalog

SOURCES.md

@@ -44,6 +44,7 @@ When a source is volatile, include a retrieval date or `last_checked` value in `
 | [openai/openai-agents-js](https://github.com/openai/openai-agents-js) | JavaScript/TypeScript agent framework | Official GitHub repository |
 | [openai/openai-agents-python](https://github.com/openai/openai-agents-python) | Python agent framework | Official GitHub repository |
 | [openai/openai-cookbook](https://github.com/openai/openai-cookbook) | Example implementation patterns | Official GitHub repository |
+| [alanhoff/agent-coordinator at fb3d646](https://github.com/alanhoff/agent-coordinator/blob/fb3d64688a9e1b7c02a114a14faac1219ec10f51/README.md) | Per-user Codex packaging, Python requirement, revisioned state, write scopes, checks, reconciliation, and Git boundary | Project-owned commit-pinned source; submitted by the maintainer; recheck later revisions |
 | [Model Context Protocol docs](https://modelcontextprotocol.io/docs/getting-started/intro) | MCP concept and terminology | Official MCP docs |
 | [github/github-mcp-server](https://github.com/github/github-mcp-server) | GitHub MCP server listing | Official GitHub repository |
 | [microsoft/playwright-mcp](https://github.com/microsoft/playwright-mcp) | Browser automation MCP listing | Official Microsoft repository |

TOOLS.md

@@ -66,6 +66,7 @@ Framework entries belong here when they help someone build Codex-adjacent tools,
 | [openai/openai-agents-js](https://github.com/openai/openai-agents-js) | Build JavaScript or TypeScript agents | Review releases, examples, and security posture before adoption |
 | [openai/openai-agents-python](https://github.com/openai/openai-agents-python) | Build Python agents | Review releases, examples, and security posture before adoption |
 | [openai/openai-cookbook](https://github.com/openai/openai-cookbook) | Find examples and implementation patterns | Examples are starting points, not production guarantees |
+| [alanhoff/agent-coordinator](https://github.com/alanhoff/agent-coordinator) | Keep long Codex jobs in a bounded work graph with explicit dependencies, revisioned local state, declared write scopes, and authored checks | Review the graph, write scopes, check commands, and any uncertain-operation reconciliation; authored checks are evidence, not proof |
 
 ## Suggest A Tool
 

data/catalog.yml

@@ -421,3 +421,18 @@ items:
     submitter_affiliation: Maintained by CodexWork from public official source.
     status: sourced
     last_checked: "2026-07-07"
+  - id: agent-coordinator
+    type: tool
+    name: alanhoff/agent-coordinator
+    url: https://github.com/alanhoff/agent-coordinator
+    source_url: https://github.com/alanhoff/agent-coordinator/blob/fb3d64688a9e1b7c02a114a14faac1219ec10f51/README.md
+    category: developer and agent frameworks
+    audience: Codex users running long or interruption-prone coding jobs
+    work_problem: Preserve explicit work dependencies, ownership boundaries, and recoverable progress across a long Codex job.
+    codex_relationship: Independent per-user Codex skill with a bundled Python 3.11+ runtime that manages revisioned local state.
+    output: Revisioned local workflow state with node status, requirements, dependencies, write scopes, acceptance criteria, checks, and reconciliation records.
+    review_boundary: Review the graph, write scopes, check commands, reconciliation records, and resulting artifacts before accepting completion.
+    limits: Requires Codex and Python 3.11+; state is local; the skill does not invoke or inspect Git; authored checks are evidence rather than proof; specialist agents are optional.
+    submitter_affiliation: Submitted by the project maintainer, Alan Hoffmeister.
+    status: sourced
+    last_checked: "2026-09-01"