This directory is the committed knowledge base for iterion’s catalog bots.
Every time a bot is dogfooded with a real run, the operator distils the outcome
into a dated bilan in docs/bot-runs/<bot>.md (named by bot directory,
not persona). The next person to launch that bot reads its file first — what it
caught, what it missed, what to change, and which engine bugs the run surfaced.
A bilan is durable, reviewable in a PR, and shows up in git log. It is one of
three complementary knowledge channels — do not confuse them:
| Channel | Where | Scope | Lifetime |
|---|---|---|---|
| Workspace memory | ~/.iterion/projects/<key>/memory/ (gitignored) — see memory-and-knowledge.md |
per-operator scratch, “what did we learn this session” | local to one machine/operator |
| Board issues | native kanban (.iterion/, gitignored) |
open tasks / findings to act on | until closed |
| Bilans (this dir) | docs/bot-runs/<bot>.md (committed) |
durable lessons the next operator must read before launching the bot | forever, in git history |
Cross-bot lessons (Goodhart, façade patterns, asymptote rules) live in workflow_authoring_pitfalls.md, not here — this directory is per-bot.
Append one dated section per run to docs/bot-runs/<bot>.md (newest first):
## YYYY-MM-DD — <short label> (run <id-prefix>)
- Status: validated | partial | failed
- Versions: bot <manifest version> · iterion <git sha>
- Method: backend(s)/model(s), budget, key --vars, flags (--merge-into, post_to_board, sandbox image)
- Result: converged? iterations, cost $, duration, where commits landed (branch/sha)
- Value: the high-value thing it actually produced (or: low value + why)
- Findings / misses: what the bot caught or missed
- Engine hardening: iterion bugs found → commits/ADRs
- Lessons for next run: what to change (vars, prompt, scanner, skill)
The run artifacts (.iterion/runs/<id>/) are gitignored, so the bilan is the
only committed trace. Regenerate the full chronological run report any time with:
iterion report --run-id <id> --output /tmp/<bot>-<id>.md
and cite the run-id in the bilan so it can be reconstructed.
Persona → bot directory, with current validation status. Add the link when the first bilan for a bot lands.
| Persona | Bot | Kind | Bilan |
|---|---|---|---|
| Nexie | whats-next |
orchestrator / board triage | whats-next.md |
| Evoly | evolve |
strategic vision + evolution proposals (per-bot memory) | evolve.md |
| Willy | whole-improve-loop |
whole-repo review-fix loop | whole-improve-loop.md |
| Billy | branch-improve-loop |
branch-scoped review-fix loop (commits) | branch-improve-loop.md |
| Featurly | feature-dev |
one-shot feature dev + review loop | feature-dev.md |
| Testy | test-coverage |
test-coverage augmentation + anti-façade review loop | test-coverage.md |
| Doki | docs-refresh |
docs↔code convergence loop | docs-refresh.md |
| Revi | review-pr |
read-only cross-family reviewer | review-pr.md |
| Revi (converse) | revi-converse |
conversational PR follow-up | not yet |
| Seki | sec-audit-source |
source SAST audit | sec-audit-source.md |
| Depsy | sec-audit-deps |
supply-chain SCA audit (scaffold) | sec-audit-deps.md |
| Renovacy | secured-renovacy |
dependency upgrade pipeline | secured-renovacy.md |
| Bmady | bmady |
BMAD multi-persona human-gated delivery | bmady.md |
| Devy | devbox-setup |
devbox.json bootstrap | devbox-setup.md |
| Adry | adr-cartograph |
ADR cartographer + completeness audit (idempotent) | adr-cartograph.md |
| Vetty | dep-update-guard |
Dependabot/Renovate PR guard (audit + align + deterministic verify) | dep-update-guard.md |
| Acci | rgaa-audit |
RGAA 4.1.2 accessibility audit (read-only) | rgaa-audit.md |
| ReArchi | adr-rechallenge |
human-gated ADR re-challenge | adr-rechallenge.md |
| Fini | feature-gap-fill |
gap-driven feature completion loop | feature-gap-fill.md |