@nlaprell/shipit 1.0.0

package/work/intent/templates/infra-change.md

@@ -0,0 +1,121 @@
+# F-###: Title
+
+## Type
+
+feature | bug | tech-debt
+
+## Status
+
+planned | active | blocked | validating | shipped | killed
+
+## Priority
+
+p0 | p1 | p2 | p3
+
+## Effort
+
+s | m | l
+
+## Release Target
+
+R1 | R2 | R3 | R4
+
+## Motivation
+
+(Why it exists, 1–3 bullets)
+
+## Resources
+
+- (Infra resources touched: e.g. DB, queue, k8s, CDN)
+- (New resources or changes to existing)
+
+## Rollout Steps
+
+- (Ordered steps: e.g. 1) Deploy new config 2) Migrate data 3) Switch traffic)
+- (Checkpoints and validation between steps)
+
+## Rollback
+
+- (How to roll back each step)
+- (Runbook or script references)
+- (Rollback tested before rollout)
+
+## Confidence
+
+Requirements: 0.0–1.0  
+Domain assumptions: 0.0–1.0
+(If either < threshold, document why proceeding or request human interrupt)
+
+## Invariants (Hard Constraints)
+
+- No single point of failure introduced
+- Rollback path tested (dry run or staging)
+- (Add infra-specific invariants)
+
+Executable (add to \_system/architecture/invariants.yml if applicable):
+
+```yaml
+invariants:
+  - rollback_tested: true
+success_metrics:
+  - zero_data_loss_on_rollback: true
+```
+
+## Acceptance (Executable)
+
+These are the ONLY criteria for "done." If it can't be checked automatically, it doesn't belong here.
+
+- [ ] Infra tests or runbooks: deployment/rollback steps validated
+- [ ] Rollback tested (staging or dry run)
+- [ ] CLI: `pnpm test` green (if app tests apply)
+- [ ] (Add infra-specific checks: e.g. health checks, capacity)
+
+## Assumptions (MUST BE EXPLICIT)
+
+- (ex: "Maintenance window is available" — confirm?)
+- (ex: "Backup is current before migration" — verify?)
+
+## Risk Level
+
+low | medium | high
+
+Risk triggers:
+
+- high: auth, data migration, permissions, infra, PII
+- medium: new dependencies, capacity, networking
+- low: docs, config-only changes
+
+## Kill Criteria (Stop Conditions)
+
+- Rollback fails or is not tested
+- Single point of failure introduced
+- Data loss risk without mitigation
+- (Add infra-specific kill criteria)
+
+## Rollback Plan
+
+REQUIRED before implementation begins. Must match "Rollback" section above.
+
+- (Step-by-step rollback; same order as Rollout in reverse where applicable)
+- (Data: down migration or restore from backup)
+- (Config: revert to previous version)
+
+## Dependencies
+
+- (Other intent IDs that must ship first)
+- (Infra team or vendor dependencies)
+- (Maintenance windows)
+
+## GitHub issue
+
+(Optional) Link to a GitHub issue. Use `#123` or `owner/repo#123`.
+
+- (e.g. `#42` or leave blank)
+
+## Do Not Repeat Check
+
+Before starting, verify this hasn't been tried before:
+
+- [ ] Checked \_system/do-not-repeat/abandoned-designs.md
+- [ ] Checked \_system/do-not-repeat/failed-experiments.md
+- [ ] No similar rejected approaches exist

package/work/intent/templates/refactor.md

@@ -0,0 +1,116 @@
+# F-###: Title
+
+## Type
+
+feature | bug | tech-debt
+
+## Status
+
+planned | active | blocked | validating | shipped | killed
+
+## Priority
+
+p0 | p1 | p2 | p3
+
+## Effort
+
+s | m | l
+
+## Release Target
+
+R1 | R2 | R3 | R4
+
+## Motivation
+
+(Why it exists, 1–3 bullets — e.g. "Reduce duplication; enable future feature X; meet new standard")
+
+## Current State
+
+- (Brief description of current structure or behavior)
+- (Pain points or constraints that motivate the refactor)
+
+## Desired State
+
+- (Target structure or behavior after refactor)
+- (Success looks like: tests unchanged or added; behavior preserved)
+
+## Scope of Change
+
+- (In scope: modules, files, APIs to touch)
+- (Out of scope: what must not change)
+- (Optional: phased steps if refactor is large)
+
+## Confidence
+
+Requirements: 0.0–1.0  
+Domain assumptions: 0.0–1.0
+(If either < threshold, document why proceeding or request human interrupt)
+
+## Invariants (Hard Constraints)
+
+- Behavior preserved: same observable behavior; same or more tests passing
+- (Add refactor-specific invariants: e.g. no API contract change)
+
+Executable (add to \_system/architecture/invariants.yml if applicable):
+
+```yaml
+invariants:
+  - behavior_preserved: true
+success_metrics:
+  - test_pass_rate: 99.5
+```
+
+## Acceptance (Executable)
+
+These are the ONLY criteria for "done." If it can't be checked automatically, it doesn't belong here.
+
+- [ ] Tests: unchanged or added; all pass (no behavior change unless explicitly in scope)
+- [ ] CLI: `pnpm test` green
+- [ ] CLI: `pnpm lint && pnpm typecheck` green
+- [ ] (Optional: snapshot or property tests updated if structure changed)
+- [ ] (Add refactor-specific checks: e.g. bundle size, coverage)
+
+## Assumptions (MUST BE EXPLICIT)
+
+- (ex: "Current tests adequately cover behavior" — add gaps?)
+- (ex: "No external callers depend on internal structure" — verify?)
+
+## Risk Level
+
+low | medium | high
+
+Risk triggers:
+
+- high: auth, money, permissions, infra, PII
+- medium: new dependencies, public API surface change
+- low: docs, internal refactors, tests, tooling
+
+## Kill Criteria (Stop Conditions)
+
+- Refactor would require breaking public contract without approval
+- Test coverage insufficient to safely refactor
+- (Add refactor-specific kill criteria)
+
+## Rollback Plan
+
+- Revert commit: clean revert possible (preferred for refactors)
+- (If phased: which phases are independently revertible)
+
+## Dependencies
+
+- (Other intent IDs that must ship first)
+- (ADRs or design docs if refactor is architectural)
+
+## GitHub issue
+
+(Optional) Link to a GitHub issue. Use `#123` or `owner/repo#123`.
+
+- (e.g. `#42` or leave blank)
+
+## Do Not Repeat Check
+
+Before starting, verify this hasn't been tried before:
+
+- [ ] Checked \_system/do-not-repeat/abandoned-designs.md
+- [ ] Checked \_system/do-not-repeat/failed-experiments.md
+- [ ] No similar rejected approaches exist