agedum 0.5.0

Drive any agent CLI from an agent-neutral source shape (AGENTS.md + .agents/skills), translating per harness at launch.

agedum

Latin agedum — "go on! / get going!"

Drive any agent CLI from an agent-neutral source shape, translating per harness at launch. You keep one set of sources; agedum renders them for whichever agent CLI you run.

• Instructions live in a root AGENTS.md (plain markdown).
• Skills live in .agents/skills/<name>/ as SKILL.md (+ optional task files, scripts, and a per-harness SKILL.<harness>.md overlay).

At launch, agedum compiles that shape to the harness's native layout in a throwaway dir, then runs your command inside a private mount namespace (bubblewrap) where the compiled files appear at their expected paths — visible only to that process, never written into your real tree or $HOME. For Claude: AGENTS.md → CLAUDE.md and .agents/skills/<name>/ → .claude/skills/<name>/ (the base SKILL.md merged with an optional SKILL.claude.md overlay).

Status: Claude harness, project + global scope, implemented. Each scope lands at its own Claude location — project → ./CLAUDE.md + ./.claude/skills/, global (~/.config/agents/AGENTS.md + ~/.agents/skills/) → ~/.claude/CLAUDE.md

• ~/.claude/skills/ (honours $CLAUDE_CONFIG_DIR). They're never merged; Claude reads both. Only those two ~/.claude paths are overlaid for the child — your ~/.claude.json auth and other settings are untouched.

kimi (--kimi) is also supported. kimi reads the project AGENTS.md natively, so agedum leaves it in place; it has no user-scope AGENTS.md, so the global AGENTS.md is injected via a transient --agent-file YAML (no --agent-file is added when there's no global scope). Skills are binds: global → ~/.kimi/skills/, project → ./.kimi/skills/ (both auto-read by kimi).

opencode (--opencode) is supported too — pure path-discovery, like Claude. The project AGENTS.md is read natively (./AGENTS.md); the global AGENTS.md binds to ~/.config/opencode/AGENTS.md; skills bind to ./.opencode/skills/ (project) and ~/.config/opencode/skills/ (global), both searched before .agents/skills/ so the overlaid copy wins. No extra flags. Linux-only; requires bwrap on PATH.

Usage

# Run a command with virtual files injected from the project + global source:
agedum --claude   -- claude --model sonnet -p "review this"
agedum --claude   -- claude                      # interactive
agedum --kimi     -- kimi -p "explain this code"
agedum --opencode -- opencode run "explain this code"

agedum --version

Everything after -- is the command, run verbatim; the context flag before -- (--claude) chooses the format. The two are decoupled, so one context can front any command. Injected paths must be gitignored — agedum refuses to overlay a git-tracked file (the namespace shares your real .git).

Documentation

Full docs at agedum.vcoeur.com:

• Source shape — the structure of AGENTS.md and .agents/skills/
• Scopes — project vs global (user) scope, and where each lands
• Harnesses — exactly what agedum does for each --<harness> command
• CLI reference and Internals — the mount-namespace launch and its safety rules

Install

pipx install agedum        # standalone CLI (once published)

Develop

make dev-install   # uv sync --all-groups
make test          # pytest
make lint          # ruff check + format --check
make run -- --version
make docs          # build the docs site (strict); docs-serve for live preview

Python ≥ 3.12, managed with uv. The version is derived from the git tag (vX.Y.Z) at build time via hatch-vcs — never committed.

Release

Tag the commit vX.Y.Z and push the tag; the release workflow builds and publishes to PyPI via OIDC trusted publishing.

License

MIT — see LICENSE.