docs(phase5): Phase 5 — Evolution Infrastructure scope#13
Merged
hoangnb24 merged 3 commits intoJun 9, 2026
Conversation
Phase 5 turns the harness from a system that verifies individual stories into one that can evolve itself. Six stories: - US-019: Machine-readable tool registry (extensible, supports non-CLI tools) - US-020: Batch story verification (completes H4) - US-021: Intervention recording schema (human corrections vs agent actions) - US-022: Context rule measurement (score-context command) - US-023: Drift detection / entropy score (audit command) - US-024: Improvement proposal pipeline (propose command, starts H5) Target: H4 partial → H4 full, H5 not achieved → H5 partial. Estimated effort: ~23-30 hours. Co-Authored-By: muscled.flumes.4r@icloud.com <hoangmrb.vn@gmail.com>
Add tool registry, batch story verification, intervention recording, context scoring, drift audit, and deterministic improvement proposals.\n\nDocument the Phase 5 operating model with audit, tool registry, improvement protocol, story packets, progress notes, schema migrations, and the durable decision for proposal rules.\n\nValidation: cargo fmt --check; cargo test --workspace; cargo clippy --workspace -- -D warnings; scripts/bin/harness-cli story verify-all; scripts/bin/harness-cli audit; scripts/bin/harness-cli propose.
Install Phase 5 reference docs, proposal decision record, and schema migrations so benchmark installs exercise the same CLI/database surface as the checked-out branch.\n\nValidation: bash -n scripts/install-harness.sh.
hoangnb24
pushed a commit
that referenced
this pull request
Jun 9, 2026
Use printf -- for post-merge changelog bullet lines so runner shells do not treat leading hyphens in format strings as options. Record the failed PR #13 workflow run and local reproduction proof in the US-025 story evidence.
hoangnb24
pushed a commit
that referenced
this pull request
Jun 9, 2026
Recover the missed post-merge maintenance for PR #13 after the workflow failed before it could write the changelog or prepare the CLI release. Record the merged PR in CHANGELOG.md, bump the Harness CLI package metadata to 0.1.9, and update the installer release tag pin to harness-cli-v0.1.9.
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
2 participants
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.
Summary
Adds
PHASE5.mddefining the next development phase for repository-harness. Phase 5 transitions the harness from verifying individual stories (Phase 4) to evolving itself — completing H4 maturity and starting H5.Six stories in dependency order:
tool register+query tools --json/--summary/--responsibilitystory verify-allcompletes H4intervention add/queryschema for human corrections vs agent actionsscore-context <trace-id>checks agent reading disciplineauditcommand with entropy score (orphaned stories, unverified records, broken tools)proposecomposes friction+interventions+drift → structured proposals (H5 begins)Key design decisions from discussion:
tool registerproposeis rule-based pattern detection (SQL grouping/counting), not LLM-generatedLink to Devin session: https://app.devin.ai/sessions/b35d6fa2be66434983cb855cda1fbda3