Codebase Audit¶
Fan three workers across your repo in parallel — dead code, API drift, and missing tests — and get independent findings without sequential bottlenecks.
Setup¶
pip install lionagi # or: uv add lionagi
# codex — requires ChatGPT Plus/Pro subscription (not an API key):
npm install -g @openai/codex
codex login
Command¶
# serialize with --max-concurrent 1 if rate-limited
li o fanout codex "Audit this repo for dead functions, API drift, and untested paths" \
-n 3 --save ./audit-out
# output:
Phase 1: Orchestrator decomposing task into 3 agent requests...
Phase 1 done (2.1s): 3 requests generated.
Phase 2: Fanning out to 3 workers: [codex/gpt-5.3-codex-spark, codex/gpt-5.3-codex-spark, codex/gpt-5.3-codex-spark]
Phase 2 done (9.4s).
Saved 3 worker results to /home/ocean/project/audit-out
════════════════════════════════════════════════════════════
Worker 1/3 [codex/gpt-5.3-codex-spark]
[...trimmed...]
How it works¶
- Workers run concurrently; results write to
./audit-out/worker_1.md…worker_3.md - Branch snapshots land in
~/.lionagi/runs/<run_id>/branches/ - Resume any worker after the run with the branch ID printed in the hints:
li agent -r <branch_id> "List the top 3 dead functions by call-site count."
Next¶
- Research synthesis — add
--with-synthesisto consolidate findings - CLI reference:
li o fanout— all flags and defaults