Agent harness distill¶
Audience: Maintainers + agents deciding what to stop loading
Success criterion (agent-facing): Less duplicative / oral context, same
judgement — gates and stems still win.
Related: stems/epistemic-layout.md, improve/oral-history.md, ADR INDEX
This is not a human-docs polish exercise. The consumer is a coding agent that overweights any prose placed in always-on paths.
Success criteria (agent consumer)¶
| Criterion | Check |
|---|---|
| Map is a table, not a chat log | capability-map ≪ 300 lines; no multi-KB cycle digests |
| One doctrine per topic | Goal B: interesting-saas-context + playbook only; antagonist report ≤1 screen |
| ADRs are on-demand | INDEX has Active set; agents not told to pre-load the ledger |
| Lore has a home | Durable thrash classes → improve/oral-history.md once |
| Stamp rule is cheap | Driver stamps Last-exercised cells only |
| Gates > prose | Counter-priors / residual / dig contracts still enforce hard truths |
| No resurrected thrash | Capability-sweep does not re-grow cycle-note archives |
If an agent still needs a deleted paragraph to act correctly, the gate or stem was wrong — restore as a stem/counter-prior/gate, not as oral dump.
Done (2026-08-10)¶
| Surface | Before | After |
|---|---|---|
.claude/commands/improve/capability-map.md |
~1700 lines (cycle digests) | ~150 lines: vocab + registry + ≤5 sweep one-liners |
docs/reference/antagonist-report-post-5-8.md |
Second full doctrine | Short pointer to interesting-saas-context |
docs/adr/INDEX.md |
Flat must-scan ledger | Active table + historical full list |
improve.md stamp rule |
Implied narrative stamps | Explicit: table only; lore → oral-history |
| Path honesty | Mixed improve/capability-map.md |
Canonical .claude/commands/improve/… |
| MCP boundary (AGENTS) | Session-era freeze language | Stateless protocol + CLI for long work |
Not bulk-deleted (still high signal): stems, counter-priors, dig contracts, closed depth menu, strategy playbooks (action steps).
Still optional later¶
| Candidate | Why wait |
|---|---|
| HM stem length | Split matrix vs judgement when editing those stems |
| AGENTS architecture table | Drift-gated; trim only with intentional gate update |
| Strategy fold (gallery vs coherence) | Real lanes; not pure duplication |
| Capability-map row count | Inventory itself is useful; only prose was thrash |
Anti-patterns this pass removes¶
- Capability-map as improve-log mirror — re-told every ship; used as “what to do next” instead of residual/policy/portfolio.
- Doctrine triple-copy — same Goal B rules in antagonist report + doctrine + playbook (agent re-decides which is canonical).
- ADR pre-load — 50+ long paragraphs of subsystem history as if curriculum.
- Session-MCP folklore as always-on fear — “MCP freezes the agent” without distinguishing long tools vs knowledge tools after stateless SDK.
Maintenance rule¶
When you learn a durable improve lesson:
- Prefer a gate or probe residual.
- Else one bullet in
improve/oral-history.md. - Else a stem / counter-prior.
- Never append another essay to capability-map.