| name | lope-execute |
| description | Execute a sprint phase-by-phase with validator-in-the-loop retry. For each phase: implement (code, deliverables, research, etc.) → AI validators review → retry on NEEDS_FIX → advance on PASS. Works with any domain (engineering, business, research). |
Lope Execute
For each phase: implement → validators review → retry on NEEDS_FIX → advance on PASS.
This skill is invoked two ways:
- Explicit: the user typed
/lope-execute <sprint_doc> and expects you to run that doc
- Implicit: the user said something like "run the auth sprint" or "execute the plan we just negotiated" and you mapped it to this skill. In that case, find the sprint doc the user is referring to (most recently created
SPRINT-*.md in the working directory is usually a safe bet — confirm with the user if ambiguous).
Two-stage review (v0.3+)
Each phase gets validated twice back-to-back:
- Spec stage — "Does the implementation match the phase's Goal and Exit Criteria?" Spec NEEDS_FIX short-circuits (no quality pass until spec is met). Spec FAIL escalates immediately.
- Quality stage — "Is the implementation well-built?" Only runs if spec passed. Catches code smell, edge cases, maintainability issues.
Set LOPE_SINGLE_STAGE=1 to revert to legacy single-pass validation.
For engineering sprints, the quality stage also includes minimality audit by default. Validators should flag unnecessary dependencies, duplicate helpers, one-implementation abstractions, broad rewrites, and symptom patches, but should not fail spec-compliant work for style alone. Disable with LOPE_MINIMALITY=off; enforce material bloat with LOPE_MINIMALITY=enforce. Business/research domains stay off unless explicitly enabled.
Steps
- Determine the sprint doc path (explicit arg OR infer from context).
- For each phase in order:
a. Implement the phase goals and meet the exit criteria
b. Run validation:
PYTHONPATH=~/.lope python3 -m lope execute "$SPRINT_PATH"
c. If spec NEEDS_FIX: apply the spec-level fixes, re-validate. Quality pass won't run until spec passes.
d. If quality NEEDS_FIX: apply the quality-level fixes, re-validate.
e. On both PASS: mark phase complete, move to next
3. After all phases: run /lope-audit for the scorecard
Rules
- Mark checkboxes in the sprint doc as you complete them
- Commit after each phase passes
- Apply ALL required fixes from validator feedback before re-validating
- Do not skip or reorder phases
- If you have doubts, negotiate with your lope teammates, not the user
Domain adaptation
Lope auto-detects the sprint's domain from the doc header:
- engineering — Validators check code quality, bugs, regressions, test coverage
- business — Validators check timeline, budget, audience targeting, success metrics
- research — Validators check methodology, sampling, validity, ethical considerations
The validator review task adapts automatically. Same /lope-execute command for every domain.
Runtime safety (v0.14.0)
Use --run-timeout for the whole command; --timeout remains a per-provider-call ceiling. Inspect the emitted request plan before large work and prefer compact evidence. Automatic shaping is bounded; use chunking only when the forecast fits --max-calls and --max-chunks. For abandoned work, run lope jobs list and preview lope jobs reap --dry-run. Never use pkill -f, killall, or process-name matching.
