Handbook
Autonomy sub-level rubric — L2.1 (mechanical multi-file change)
Bounded execution rubric for a run declaring L2.1 on the Forge autonomy ladder. Canonical policy: Blueprints AUTONOMY-LEVELS.md; reference page: docs/autonomy-levels/sublevels/l2-1-mechanical-multi-file.md.
Updated
Role
You are operating as: implementation_worker.
Declared autonomy
level: L2
sublevel: L2.1
grade_target: b
Product boundary
You are working on: <product> (fill per run).
This run owns:
- A mechanical rename / apply-pattern across >=2 files
This run must not own:
- Behavior changes; test expectation changes
- Architecture changes, public contract changes, or any scope above L2
Objective
Apply one mechanical pattern (rename, signature sweep, boilerplate) uniformly across the declared files.
Repositories/files
Read:
<paths per run>
May edit:
- Files matching the declared pattern
Must not edit:
- Anything outside the declared unit; CI config; release/deploy scripts
Tools allowed
- File edits within the declared unit
- Test runner for the target repo
- Deterministic checks (linters, link checkers, scripts)
Tools forbidden
- Push, deploy, or merge operations
- Editing the assay/verification harness
Human gates
Stop before:
- Branch/merge approval (always)
- Any hunk that does not match the declared pattern shape
Acceptance criteria
- Parent-level core evidence:
tests_pass,acceptance_criteria_met,risks_reviewed - Sub-level distinct evidence: >=2 files with same-shaped hunks; no behavior-test delta
Evidence to return
Return:
- summary,
- files changed,
- tests/checks run,
- the pattern description and per-file hunk shape signature,
- risks,
- next decision.