Conversation
Adds pip install --pre --upgrade specsmith + migrate-project + audit + sync before every session. Agents now always run system-installed specsmith. Co-Authored-By: Oz <oz-agent@warp.dev>
…, query API, metrics) Brings all Phase 2 work (PR #10) into the develop branch. Conflict resolution: kept develop's Session Bootstrap section; updated test count to 192. Co-Authored-By: Oz <oz-agent@warp.dev>
Root cause fixed: governance-mode was 'strict' causing specsmith to read both docs/requirements/chronomemory.yml AND docs/REQUIREMENTS.md as independent sources, creating 14 duplicate IDs. Switched to yaml mode (single canonical source = YAML; MD is derived via specsmith sync). Phase 1 (REQ-CM-001..014): YAML updated to match current implementation with platform, boundary, status=implemented, correct component paths. Phase 2 added (REQ-CM-015..021): REQ-CM-015 Typed Dependency Graph (DepGraph) REQ-CM-016 Epistemic Rollback and Cascade Propagation REQ-CM-017 Context Pack Compiler REQ-CM-018 Extended Query API - ESDB Spec s23 (18 functions) REQ-CM-019 Token Metrics Tracking REQ-CM-020 Skill System - First-Class Skill Records REQ-CM-021 ChronoStore Phase 2 Convenience Methods TEST-CM-015..021 added to docs/tests/chronomemory.yml covering all 192 test cases across test_deps, test_rollback, test_context_pack, test_query_ext, and test_metrics. ARCHITECTURE.md updated: added module docs for deps.py, rollback.py, context_pack.py, query.py, metrics.py; Rust crate moved to Phase 3. Validate: 21 unique REQ IDs, 4/4 std checks clean, 0 strict errors. Audit: 28/28 checks, 21 REQs covered, Phase Release 100%. Co-Authored-By: Oz <oz-agent@warp.dev>
…ibrium
Eliminated 7 logic knots (negation conflicts) by rephrasing MUST NOT
language in three requirement descriptions:
REQ-CM-002: 'must not be the default format' ->
'NDJSON is the required default; binary is Phase 2 only'
REQ-CM-008: 'must not be conflated' ->
'they are distinct concepts and must remain separate'
REQ-CM-010: 'must not vary between consumers' ->
'are invariant across all consumers and versions'
Stress-test results (before -> after):
Logic knots: 7 -> 0
Failure modes: 24 -> 3 (all LOW severity, no critical)
Equilibrium: NO -> YES
3 remaining LOW suggestions: decompose REQ-CM-015/016/020 (compound
claims). These are stylistic — each REQ already has its own test spec.
Final state: validate 4/4 std, audit 28/28, stress-test equilibrium YES,
192 tests passing, ruff clean, mypy clean.
Co-Authored-By: Oz <oz-agent@warp.dev>
Decomposed 3 compound Phase 2 requirements into focused single-proposition REQs to achieve 0 failure modes in stress-test: REQ-CM-015: Typed Dependency Graph - Core Interface and BFS Traversal REQ-CM-022: DepGraph WAL Persistence and Replay (split from 015) REQ-CM-016: Epistemic Rollback - Cascade Propagation Logic (simplified) REQ-CM-023: Rollback WAL Event and ChronoStore Integration (split from 016) REQ-CM-020: Skill System - find_skills Keyword Matching (simplified) REQ-CM-024: Skill System - record_skill_run WAL Record (split from 020) Added TEST-CM-022, TEST-CM-023, TEST-CM-024 linked to existing test files. Stress-test results: 0 failure modes, 0 logic knots, Equilibrium YES. Audit: 28/28, 24 REQs covered, Phase Release 100%. Validate: 24 unique IDs, 4/4 std, 0 strict errors. Tests: 192 passing. ruff: clean. mypy: clean. Co-Authored-By: Oz <oz-agent@warp.dev>
what_is_known() was leaking kind=edge, rollback_event, token_metric, and
skill_run records into RAG query results. These are infrastructure
bookkeeping records, not agent-facing beliefs.
Fix: when no kind= filter is specified, filter out _INFRA_KINDS before
returning. The kind= parameter still works as a passthrough to retrieve
infra records explicitly when needed.
Impact for consumer repos:
specsmith: query(rag_filter=True) unaffected (direct store call)
chronoagent: what_is_known() now returns only belief records, no graph
metadata leaking into LLM context
ctt-neural: what_is_known() unaffected (uses kind='fact' filter)
All 192 tests pass. No existing tests broken.
Co-Authored-By: Oz <oz-agent@warp.dev>
Co-Authored-By: Oz <oz-agent@warp.dev>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Brings main up to date with all develop improvements since PR #10:
Bugfix:
Governance:
All checks: 192 tests, ruff clean, mypy clean, audit 28/28, stress-test equilibrium YES.
Co-Authored-By: Oz oz-agent@warp.dev