Apothem vs raw Gemini CLI config
Hand-maintaining Gemini CLI's GEMINI.md rules file versus rendering it from Apothem's shared profile.
Gemini CLI reads a single GEMINI.md rules file. Apothem's gemini-cli adapter
renders it from the shared profile. This page contrasts hand-maintaining that
config against managing it with Apothem.
Effort comparison
| Task | Raw hand-config | Apothem-managed |
|---|---|---|
| Author rules | Hand-edit one GEMINI.md | Rendered: project-wide guidance in GEMINI.md; rules in .gemini/.apothem/support/ |
| Keep rules aligned with other harnesses | Manual, per tool | Automatic on apothem update |
| Install / remove the config | Write and delete GEMINI.md by hand | apothem install / apothem uninstall |
| Verify the config is correct | Manual inspection | apothem verify --harness gemini-cli --project <path> |
What Apothem manages
The gemini-cli adapter writes a sentinel-delimited Apothem block into
<project>/GEMINI.md that carries project-wide guidance and points to Apothem's
material. Rules, templates and hooks land as support files under
<project>/.gemini/.apothem/support/, and commands, skills and agents land under
<project>/.gemini/. Gemini CLI documents a hooks object in settings.json,
which stays operator-owned, so the adapter registers no hook and guardrail intent
is expressed in the rule prose rather than a tool-call interception point. Every
change to the shared rules re-renders GEMINI.md on the next apothem update.
Migration path
- Run
apothem profile initto create the shared profile if you have not yet. - Move the rules from your existing
GEMINI.mdinto the shared corpus. - Run
apothem install --harness gemini-cli --project <path>to render and writeGEMINI.md. - Run
apothem verify --harness gemini-cli --project <path>to confirm the config landed. - Use
apothem update --harness gemini-cli --project <path>for every later change.