Comparing changes

@bug

…or v5 (#70) * chore(workflow): refine agents, skills, commands, and ADR structure for v5 - Consolidate ADR files into single adr.md (drop adr-NNN-<title>.md pattern) - Add test-coverage and test-build tasks; fix test-fast/test flags - Clarify design priority chain with full complexity ladder - Remove self-selection language from agent instructions - Add refactor and design-patterns skills to reviewer agent - Add BASELINED guard to feature-selection skill - Fix scope skill: drop silent pre-mortem requirement from Session 1 - Align function/class line-count wording to code-lines-only * chore(workflow): document model v6 — 2-stage discovery, PO-owned feature moves, unified session protocol - Split docs/discovery.md into three append-only files: docs/discovery_journal.md (raw Q&A), docs/discovery.md (synthesis), docs/architecture.md (architectural decisions) - Remove docs/features/discovery.md (superseded) - Replace Phase 1/Phase 2 model with 2-stage model: Stage 1 Discovery (unified iterative sessions) + Stage 2 Specification (stories + criteria) - PO is sole owner of all .feature file moves; SE and reviewer never move or edit .feature files; explicit escalation when no in-progress feature - Add bug handling protocol: PO adds @bug @id, SE writes both @id test and @given Hypothesis property test - Add real-time split rule: >2 concerns or >8 Examples splits immediately within the same session - Add session status markers to discovery_journal.md: IN-PROGRESS/COMPLETE - Update all agent files, skills, workflow docs, and research docs to match

- scope/SKILL.md: add mandatory PO Self-Declaration block (11 claims covering INVEST I/V/S/T, observable outcomes, no impl details, entity coverage, distinct examples, unique IDs, pre-mortem, scope boundary) after Stage 2 Step B review checklist; every DISAGREE is a hard blocker before commit - session-workflow/SKILL.md: add Step 1 Stage 2 Criteria Self-Declaration TODO format section with full template; add Rule 9 enforcing declaration before criteria commit - product-owner.md: note Self-Declaration requirement in Step 1 routing table - verify/SKILL.md: add Reviewer Stance Declaration section (5 claims: adversarial mindset, manual trace, boundary check, semantic read, independence) to report template between SE Self-Declaration Audit and Decision; DISAGREE allowed with explanation, unexplained DISAGREE = REJECTED

…72)

…n AGENTS.md (#73)

…mmunication (#74)

…tooling concern (#75)

* fix(verify): correct stale Self-Declaration location references SE Self-Declaration is communicated verbally in the handoff message, not written to TODO.md. Three references in verify/SKILL.md still pointed to TODO.md from the v4.1 era when it was a written block. * refactor(reviewer): slim reviewer.md and verify skill to reduce token load - reviewer.md: remove default hypothesis, After APPROVED block, and Available Skills list — all duplicated from verify/SKILL.md which is always loaded in the same session (ai-agents.md #24, #26) - verify/SKILL.md v4.0: - drop Standards Summary table (pure restatement of inline content) - drop 'When to Use' section (no information beyond skill name) - OC size/nesting rules (5d rules 1 and 7) cross-reference 6b thresholds instead of repeating values - replace 5d OC table with design-patterns skill reference (on-demand loading per ai-agents.md #23); commitment-device structure preserved inside that skill - move Self-Declaration Audit to Step 5 (before code review) — if SE flagged a DISAGREE, reviewer saves all of code review before rejecting - DISAGREE handling: triage gate instead of hard reject — accepts genuine out-of-SE-control constraints with a note and suggested alternative; rejects weak/unjustified deviations with specific fix - renumber code review sub-sections 5a-5g → 6a-6g * chore(release): bump version to v6.1.20260419 - Contextual Ambystoma

…tegration (#77) * fix(verify): correct stale Self-Declaration location references SE Self-Declaration is communicated verbally in the handoff message, not written to TODO.md. Three references in verify/SKILL.md still pointed to TODO.md from the v4.1 era when it was a written block. * refactor(reviewer): slim reviewer.md and verify skill to reduce token load - reviewer.md: remove default hypothesis, After APPROVED block, and Available Skills list — all duplicated from verify/SKILL.md which is always loaded in the same session (ai-agents.md #24, #26) - verify/SKILL.md v4.0: - drop Standards Summary table (pure restatement of inline content) - drop 'When to Use' section (no information beyond skill name) - OC size/nesting rules (5d rules 1 and 7) cross-reference 6b thresholds instead of repeating values - replace 5d OC table with design-patterns skill reference (on-demand loading per ai-agents.md #23); commitment-device structure preserved inside that skill - move Self-Declaration Audit to Step 5 (before code review) — if SE flagged a DISAGREE, reviewer saves all of code review before rejecting - DISAGREE handling: triage gate instead of hard reject — accepts genuine out-of-SE-control constraints with a note and suggested alternative; rejects weak/unjustified deviations with specific fix - renumber code review sub-sections 5a-5g → 6a-6g * feat(release): integrate living-docs into git-release process

@id

#78) * chore(workflow): integrate pytest-beehave and clean up @id conventions - Add pytest-beehave[html]>=3.0 to dev deps; configure [tool.beehave] - Fix test-build task: --cov=pytest_beehave -> --cov=app - Remove manual deprecated skip hook from conftest.py (beehave owns it) - Naming: feature-stem for .feature paths, feature_slug for test dirs - @id tags now auto-assigned on first pytest run; remove all manual generation instructions - Test stubs auto-generated at Step 2 end via test-fast; remove manual stub section - Add pytest-beehave to README tooling table; add Why section and auto-gen stub example - Update product-owner.md, AGENTS.md, and all affected skills accordingly * chore(workflow): number self-declaration items and add completeness enforcement - implementation/SKILL.md: number all 25 items 1-25; add count reminder comment - verify/SKILL.md: add completeness hard gate (count must be 25, sequence must be gapless); expand report table from 21 to 25 numbered rows matching implementation template exactly

)

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Comparing changes

Open a pull request

Commits on Apr 19, 2026

This comparison is taking too long to generate.

Uh oh!