agentleFS
Sign inStart free
docxologydocxology/codomyrmex/SKILL.md

Codomyrmex

Skill10 starsChanged 38 days agoMITView on GitHub

Full-spectrum coding workspace skill providing 627 production MCP `@mcp_tool` lines across 130 top-level modules; the complete locked dependency profile currently exposes 612 tools in the merged runtime manifest. USE WHEN user says 'verify codomyrmex', 'codomyrmexVerify', 'audit codomyrmex', 'trust codomyrmex', 'codomyrmexTrust', 'trust tools', 'enable destructive tools', 'check pai status', 'codomyrmex tools', 'codomyrmex analyze', 'codomyrmex search', 'codomyrmex memory', 'codomyrmex docs', 'codomyrmex status', 'codomyrmex git', 'codomyrmex security', 'codomyrmex ai', 'codomyrmex code', 'codomyrmex data', 'codomyrmex deploy', 'codomyrmex test', or uses any 'codomyrmex' automation tools.

The file

Create a free account to mount
Files
SKILL.md

Codomyrmex Skill for PAI

Package version: v1.3.0 (tracks pyproject.toml) | Type: Infrastructure Skill | MCP: codomyrmex-mcp-server | Skills: 15 | Tools: 627 production @mcp_tool lines; 612 merged runtime tools in the complete locked dependency profile (inventory)

Description

Full-spectrum coding workspace skill providing 627 production @mcp_tool lines across 130 top-level modules for AI-assisted development, code analysis, testing, documentation generation, and workflow automation. In the complete locked dependency profile, the merged runtime manifest currently exposes 612 tools because decorator lines and registered runtime entries are distinct measurements.

Installation

# codomyrmex is not published on PyPI; install it from a checkout
uv sync

Quick Start

# MCP server (for PAI agent consumption)
from codomyrmex.agents.pai.mcp_bridge import create_codomyrmex_mcp_server
server = create_codomyrmex_mcp_server()
server.run()

# Direct Python calls (no MCP overhead)
from codomyrmex.agents.pai import verify_capabilities, trust_all, trusted_call_tool

verify_capabilities()   # Step 1: Audit & promote safe tools
trust_all()             # Step 2: Grant full execution trust
result = trusted_call_tool("codomyrmex.write_file", path="x.py", content="...")

Codex Access Probe

Codex sessions can inspect the repo-native agentic surface without mutating the workspace or launching agents:

uv run python scripts/agents/codex_access.py --json
uv run python scripts/agents/codex_access.py --check --json
uv run python scripts/agents/improve_src.py --dry-run --limit 2 --json

The same read-only payloads are exposed as MCP tools:

  • codomyrmex.codex_access_status - MCP, skills, trust, Hermes, and Codex readiness.
  • codomyrmex.codex_dispatch_catalog - multiagent dispatch paths classified as read_only, dry_run, or side_effectful.

Skill Domains (15)

SkillDomainKey Triggers
codomyrmexVerifyCapability Audit"verify codomyrmex", "audit tools"
codomyrmexTrustTrust Management"trust codomyrmex", "enable destructive"
codomyrmexAnalyzeCode Analysis"analyze project", "code review"
codomyrmexMemoryMemory"add to memory", "remember this"
codomyrmexSearchSearch"search codebase", "grep pattern"
codomyrmexDocsDocumentation"get module docs", "module readme"
codomyrmexStatusStatus Dashboard"system status", "health check"
CodomyrmexGitVersion Control"git analysis", "commit timeline"
CodomyrmexSecuritySecurity & Crypto"security scan", "crypto key"
CodomyrmexAIAI & Agents"reasoning trace", "thinking agent"
CodomyrmexCodeCode Execution"execute code", "sandbox code"
CodomyrmexDataData & Visualization"bar chart", "fuzzy search"
CodomyrmexDeployInfrastructure"docker build", "list instances"
CodomyrmexTestTesting"run tests", "benchmark"

Tools Summary (627 source decorators; 612 merged runtime tools)

Most tools are read-only; a small set is destructive (require /codomyrmexTrust). Run /codomyrmexVerify for the live safe/destructive split.

  • Destructive: write_file, run_command, run_tests, call_module_function

The source inventory contains 627 production decorator lines across 151 mcp_tools.py files and 130 top-level modules. Runtime registration is a separate measured surface: the merged manifest exposes 612 tools in the complete locked dependency profile, while the standalone full launcher and readonly HTTP profile expose narrower configured profiles.

Resources

  • codomyrmex://modules — Full module inventory (JSON)
  • codomyrmex://status — System + PAI status (JSON)

Algorithm Phase Mapping

PhasePrimary Tools
OBSERVElist_modules, module_info, health_check, dependency_tree
THINKanalyze_python, search_codebase, think, query_knowledge_base
PLANread_file, json_query, analyze_workflow_dependencies
BUILDwrite_file, generate_module_docs, code_execute
EXECUTErun_command, run_tests, execute_agent, container_build
VERIFYgit_diff, checksum_file, solve_model, audit_code_security
LEARNpai_awareness, memory_put, emit_event

Trust Levels

LevelAccessHow
UNTRUSTEDNoneDefault state
VERIFIEDRead-only tools/codomyrmexVerify
TRUSTEDAll registered dynamic tools (627 source @mcp_tool lines; 612 merged runtime tools in the complete locked dependency profile; see inventory) plus PAI static proxy tools/codomyrmexTrust

Repository

Source: src/codomyrmex/agents/pai/ — MCP bridge, trust gateway, PAI bridge

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

Check your setup, free