Aider loads a conventions file into context and keeps it there for the whole
session — that is what the file is for, and the docs say to load it with
--read so prompt caching can hold it. This integration concatenated every
agent body into it. At 279 agents that is 3,816,372 characters, roughly a
million tokens. No model takes that. Anyone who ran
./scripts/install.sh --tool aider
got a CONVENTIONS.md that either blows the context window on the first turn
or bills for a million tokens trying.
CONVENTIONS.md is now the roster index it was described as: one entry per
agent with the name, the description, the division, and the path to the
agent file. 96,823 characters, down from 3.8 million. The header explains
how to pull a single agent's full instructions into the session:
/read-only /path/to/agency-agents/engineering/engineering-frontend-developer.md
Naming an agent in a prompt still works the way it did — the description is
what the model needed for that, and it is still there.
test-convert-outputs.sh now holds the index to being an index: it fails if
CONVENTIONS.md grows past 250,000 characters, if it does not list exactly
one path per roster agent, or if any path it prints does not resolve. The
existing round-trip check on the accumulated file still covers the names and
descriptions.
- Antigravity installs to ~/.gemini/config/skills/ per antigravity/README.md,
convert.sh, and install.sh — the index was the sole outlier still saying
~/.gemini/antigravity/skills/
- Mistral Vibe is first-class in convert.sh and install.sh and has its own
integrations/vibe/README.md, but was missing from the Supported Tools index
(The originality-gate fix originally in this PR was superseded by #659.)
Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
Migrates Gemini CLI to native subagents (~/.gemini/agents/) + quotes zk-steward description. Rebased from #472; e2e-verified with real gemini v0.43.0. Closes#473.
Co-Authored-By: Tomo Wang <tomo_wang@163.com>
Adds Codex as a conversion/install target: each agent → `~/.codex/agents/<slug>.toml` with the three required Codex fields (name, description, developer_instructions).
Validated: all 184 agents generate valid, parseable TOML (incl. 21k-char agents with embedded code blocks) via the PR's TOML basic-string escaper. Matches OpenAI's documented custom-agent schema.
Thanks @yunuskilicdev.
Add complete support for Kimi Code CLI agent format.
- Add convert_kimi() function to generate YAML agent specs
- Add install_kimi() function to install agents to ~/.config/kimi/agents/
- Add Kimi to tool detection and installer UI
- Add integrations/kimi/ directory (generated files gitignored)
- Update integrations/README.md with Kimi documentation
- Add generated agent directories to .gitignore
Users can generate agents with: ./scripts/convert.sh --tool kimi