0030790283
feat(skills): encode evidence-driven anti-patterns, unify flywheel metrics, add normalization defect detection
...
- Create global ~/.agents/MEMORY.md with 7 universal principles and top patterns
- Unify sigma/rho/delta definitions across skills/docs/code (code is truth)
- Add normalization defect scan to Athena defrag (placeholder, stacked frontmatter, bundled, dupes)
- Add anti-pattern pre-flight to /plan (dead infra, brainstorm-as-plan, 40% context, commit cadence)
- Expand pre-mortem from top-5 to top-8 FAIL patterns (dead infra activation, rollback maps, four-surface closure)
- Add 4 new forge uncaptured lesson categories (normalization defects, promotion pipeline, taxonomy, stale contradictions)
- Add normalization defect check to post-mortem four-surface closure
- Fix pre-existing broken link in flywheel-diagnostics.md (regex in code block)
- Sync all changes to skills-codex with Codex-specific adaptations preserved
2026-03-30 15:27:49 -04:00
0e62657856
docs(flywheel): close prevention-ratchet validation loop
2026-03-10 07:44:59 -04:00
899c0eaea5
feat: add finding registry prevention ratchet
2026-03-09 16:44:54 -04:00
4abd2a69be
docs(context): sweep remaining ao inject references across docs/tests/scripts
...
Wave 3 of context-enforcement-phase3: replace ao inject workflow
guidance with ao lookup across documentation, test fixtures, and
scripts. Remaining 32 references are legitimate (inject command
implementation, CHANGELOG history, CLI reference tables, test
fixtures for command detection).
Files updated:
- 11 core docs (context-packet, seed-definition, leverage-points, etc.)
- 7 workflow diagram docs (brief, system-map, cli-skills-map, HOOKS, etc.)
- 4 scripts (proof-run, test-flywheel, check-skill-flag-refs, etc.)
- 3 shell tests (proof-run e2e, token-budgets, no-compaction)
- 1 Go test (json_validity_test.go)
- Production code: metrics_nudge.go, metrics_flywheel.go, hooks.go
2026-03-07 11:59:18 -05:00
4fd2512c87
refactor: flatten CLI namespace and liquidate debt
...
Remove 5 namespace parent commands (know, work, quality, settings, start)
and register all 49 leaf commands directly on rootCmd. Flat namespace is
simpler for users; GroupID-based help provides visual grouping without
requiring namespace tokens.
Namespace flattening (184 files):
- Delete know.go, work.go, quality.go, settings.go, start.go
- Invert deprecatedCommands map (old namespace paths → flat paths)
- Update 127 consumer files (hooks, skills, docs, scripts, tests)
Debt liquidation (7 items from post-mortem):
- Fix 62 broken smoke test invocations
- Delete 118 LOC dead deprecatedAlias code
- Widen doctor scan to docs/, scripts/, and subdirectories
- Expand expectedCmds test to all 49 commands (floor 45)
- Add bidirectional expectedCmds sync check
- Create sweep-namespace-references.sh automation
- Fix EXPECTED_COMMANDS array to list flat commands
- Escape sed variables in sweep script
All Go tests pass (36s). All conformance checks pass.
2026-02-26 09:27:38 -05:00
a539546d33
codex installer: ship no-clone install flow and docs sync
2026-02-26 06:24:21 -05:00
bdf112ba7f
feat: align codex skill docs and hook command mappings
2026-02-26 05:48:41 -05:00
c665343d2c
fix: resolve 32 vibe findings (2 critical, 7 high, 12 medium, 11 low)
...
Fix all findings from the v2.17.0..HEAD vibe review across 12 source files:
Critical: panic guards for negative budget in truncateToCharBudget and
dead prefix slice in generateArtifactID. High: bead override in batch
extract, json file counting in metrics, constraint index.json exclusion,
complexity reduction in curate status/verify (38→<25), flexible hook
script count assertion. Medium: truncate panic guards, error logging for
silent failures, null→[] JSON output, goals auto-detect (GOALS.md then
GOALS.yaml), raw var→getter usage, truncate-before-lock race fix. Low:
scanner.Err() checks, os.Stdout→cmd.OutOrStdout(), dry-run output
differentiation.
Docs: 5 missing INDEX.md concept links, curation-pipeline v1 status
callout. Tests: new TestSeed_DryRun_JSON. Regen: COMMANDS.md, embedded
hooks synced. All gates pass: build, vet, test, gocyclo, heal, doc-gate.
2026-02-24 16:51:26 -05:00
ad12717607
feat(seed): Wave 3 — Meadows leverage points documentation
...
The Seed epic (ag-poz) Wave 3: theoretical foundation made explicit.
- docs/leverage-points.md: All 12 Meadows leverage points mapped to AgentOps
architecture, dK/dt equation mapping, #5→#4 emergence insight, 6th paradigm
shift (designed→evolved), ao metrics health ↔ LP #7-#8 connection, gap analysis
Full test suite passes (20 packages, 0 failures).
Closes: ag-poz.9, ag-poz.10
2026-02-24 12:40:40 -05:00
Boden Fuller