create-quiver 0.15.2 → 0.15.4

package/specs/quiver-v41-cli-i18n-ai-lifecycle/EVIDENCE_REPORT.md

@@ -0,0 +1,90 @@
+# Evidence Report - Quiver v41 CLI i18n AI Lifecycle
+
+## slice-00-ai-i18n-foundation
+
+- Created this spec package and slice handoffs.
+- Runtime implementation intentionally not started.
+
+## slice-01-ai-run-status-resume
+
+- Routed `ai run create` through the shared language option and localized current `ai run` lifecycle wrapper errors.
+- Kept `ai run close`, `ai status`, `ai resume`, and `ai approvals` human output localized while preserving run ids, phases, statuses, commands, paths, and approval candidate versions.
+- Added Spanish regression coverage for `ai run create`, `ai run close`, and unsupported `ai run watch` wrapper errors.
+- Recorded that full `ai run watch` runtime is not present yet and remains owned by `specs/quiver-v36-ai-run-watch-portable`.
+
+Validation:
+
+- PASS `node --test tests/commands/ai-run-state.test.js`
+- PASS `node --test tests/lib/i18n-catalog.test.js`
+- PASS `git diff --check`
+
+## slice-02-ai-agent-models
+
+- Routed `ai agent` and `ai models` through the shared language option.
+- Localized agent profile reports, dry-runs, doctor/repair headings, selector prompts, actionable wrapper labels, and model catalog human output.
+- Preserved provider ids, model ids, profile JSON, model catalog JSON, and command snippets.
+
+Validation:
+
+- PASS `node --test tests/commands/ai-agent.test.js`
+- PASS `node --test tests/commands/ai-models.test.js`
+- PASS `node --test tests/lib/i18n-catalog.test.js`
+- PASS `git diff --check`
+
+## slice-03-ai-planner-approval-review
+
+- Routed `ai plan`, `ai review-plan`, `ai repair-plan`, `ai revise`, and `ai approve` through the shared language option.
+- Localized planner dry-runs, spec-generation wrappers, review/repair wrappers, approval output, progress checks, selector prompts, and actionable wrapper labels.
+- Added regression coverage proving Spanish wrappers work while provider prompt bodies, commands, phase names, version ids, draft paths, and generated spec ids remain exact.
+
+Validation:
+
+- PASS `node --test tests/commands/ai-plan-spec-phase.test.js`
+- PASS `node --test tests/commands/flow.test.js`
+- PASS `node --test tests/lib/i18n-catalog.test.js`
+- PASS `git diff --check`
+
+## slice-04-ai-execution-pr
+
+- Routed `ai execute-plan` and `ai pr` through the shared language option.
+- Localized execute-plan dry-runs, progress summaries, PR reports, and PR progress wrappers.
+- Reused existing execute-slice i18n coverage and added Spanish regression coverage for execute-plan and PR dry-runs.
+- Preserved git/gh command strings, branch names, PR titles, body paths, validation commands, JSON output, and SSH alias handling.
+
+Validation:
+
+- PASS `node --test tests/commands/ai-execute-slice.test.js`
+- PASS `node --test tests/commands/ai-execute-plan.test.js`
+- PASS `node --test tests/commands/ai-pr.test.js`
+- PASS `node --test tests/lib/i18n-catalog.test.js`
+- PASS `git diff --check`
+
+## slice-06-ai-prepare-context-progress-i18n-fix
+
+- Fixed mixed-language live progress in `ai onboard` and `ai prepare-context --with-planner`.
+- Added `en` and `es` catalog keys for base-doc reading and structure detection progress checks.
+- Updated progress tests to assert English defaults and localized approval selector hints after the planner localization work.
+- Preserved provider prompt bodies, JSON surfaces, command snippets, provider ids, and model ids.
+
+Validation:
+
+- PASS `node --test tests/commands/ai-plan.test.js tests/commands/ai-prepare-context-planner.test.js tests/commands/ai-review-plan.test.js tests/lib/i18n-catalog.test.js`
+- PASS `node --test tests/**/*.test.js` (`599/599`)
+- PASS `git diff --check`
+
+## slice-05-ai-tests-smokes
+
+- Closed v41 with full local test, package, smoke, spec, slice, and whitespace validation.
+- Confirmed provider-backed flows remain covered through injected provider runners, `--dry-run`, and `--print-prompt` tests; no live provider credentials are required.
+- Confirmed JSON/JSONL stability through the full CLI suite and existing machine-output tests.
+- Confirmed the package tarball smoke creates `create-quiver-0.15.3.tgz`.
+
+Validation:
+
+- PASS `node --test tests/**/*.test.js` (`599/599`)
+- PASS `npm run package:quiver` (`create-quiver-0.15.3.tgz`)
+- PASS `npm run smoke:create-quiver`
+- PASS `node bin/create-quiver.js spec validate specs/quiver-v41-cli-i18n-ai-lifecycle --strict`
+- PASS `node bin/create-quiver.js check-slice specs/quiver-v41-cli-i18n-ai-lifecycle/slices/slice-05-ai-tests-smokes/slice.json --local`
+- PASS `node bin/create-quiver.js check-slice specs/quiver-v41-cli-i18n-ai-lifecycle/slices/slice-06-ai-prepare-context-progress-i18n-fix/slice.json --local`
+- PASS `git diff --check`

package/specs/quiver-v41-cli-i18n-ai-lifecycle/EXECUTION_PLAN.md

@@ -0,0 +1,33 @@
+# Execution Plan - Quiver v41 CLI i18n AI Lifecycle
+
+## Execution Rules
+
+- Start only after v37 foundation is complete.
+- Localize Quiver wrapper messages only.
+- Preserve provider prompts, JSON, JSONL, and artifacts.
+- Do not change credential or provider access behavior.
+
+## Suggested Order
+
+1. `slice-00-ai-i18n-foundation`
+2. `slice-01-ai-run-status-resume`
+3. `slice-02-ai-agent-models`
+4. `slice-03-ai-planner-approval-review`
+5. `slice-04-ai-execution-pr`
+6. `slice-05-ai-tests-smokes`
+
+## Parallelization Guidance
+
+- `slice-02` and `slice-03` may overlap only after `slice-01` if `src/create-quiver/commands/ai.js` ownership is split carefully.
+- `slice-04` should wait for planner/approval behavior to stabilize.
+- `slice-05` runs last.
+
+## Final Validation
+
+```bash
+node --test tests/**/*.test.js
+git diff --check
+npx create-quiver spec validate specs/quiver-v41-cli-i18n-ai-lifecycle --strict
+npm run package:quiver
+npm run smoke:create-quiver
+```

package/specs/quiver-v41-cli-i18n-ai-lifecycle/SPEC.md

@@ -0,0 +1,57 @@
+# Quiver v41 - CLI i18n AI Lifecycle
+
+**Date:** 2026-05-28
+**Status:** Completed
+**Source:** Continuation of the approved CLI i18n program.
+
+## Problem
+
+AI lifecycle commands contain long-running progress, approvals, provider errors, and PR flows. These messages must be localized without changing provider prompts, stable artifacts, or automation contracts.
+
+## Objective
+
+Localize AI lifecycle command output in `en` and `es`, including planning, approvals, agents, execution, run status, and PR creation wrappers.
+
+## Scope
+
+### Included
+
+- `ai run create`, `ai run watch`, `ai status`, `ai resume`.
+- `ai agent set`, `ai agent doctor`, `ai agent repair`, `ai models list`.
+- `ai plan`, `ai revise`, `ai review-plan`, `ai repair-plan`, `ai approve`, `ai approvals`.
+- `ai execute-slice`, `ai execute-plan`, and `ai pr`.
+- Human progress, loaders, review/interactive wrapper messages, no-TTY/CI, dry-run, `--print-prompt`, and supported `--json`.
+
+### Excluded
+
+- Translating provider prompts or model-generated content unless a later spec explicitly defines that behavior.
+- Changing provider adapters, credentials, or model access semantics.
+
+## Acceptance Criteria
+
+1. Human output for included commands supports `en` and `es`.
+2. Provider prompts, prompt files, run ids, provider names, model ids, and command snippets remain exact.
+3. `--print-prompt` output remains an exact provider prompt surface and is not polluted by localized decoration.
+4. JSON/JSONL outputs remain parseable and schema-stable.
+5. CI/no-TTY never blocks for localized prompts.
+6. Errors from provider execution preserve actionable context without leaking secrets.
+7. Tests cover normal, dry-run, review, approval, failure, no-TTY, and JSON modes where applicable.
+
+## Slice Roadmap
+
+| Slice | Title | Status | Dependencies |
+|---|---|---|---|
+| slice-00 | AI i18n foundation | completed | none |
+| slice-01 | AI run status and resume | completed | v37 complete |
+| slice-02 | AI agents and models | completed | slice-01 |
+| slice-03 | AI planner, approvals, and review | completed | slice-01 |
+| slice-04 | AI execution and PR | completed | slice-03 |
+| slice-05 | AI tests and smokes | completed | slice-02, slice-04, slice-06 |
+| slice-06 | AI prepare-context progress i18n fix | completed | slice-03 |
+
+## Guardrails
+
+- Localize Quiver wrapper UX, not provider-owned content.
+- Keep JSONL watcher output pure.
+- Keep secrets redaction behavior unchanged or stricter.
+- Do not translate model/provider identifiers.

package/specs/quiver-v41-cli-i18n-ai-lifecycle/STATUS.md

@@ -0,0 +1,26 @@
+# Status - Quiver v41 CLI i18n AI Lifecycle
+
+**Overall status:** Completed
+**Created:** 2026-05-28
+**Completed:** 2026-05-29
+**Current slice:** none
+
+## Summary
+
+This spec localizes AI lifecycle command output after the v37 foundation is complete.
+
+## Slice Status
+
+| Slice | Status | Notes |
+|---|---|---|
+| slice-00-ai-i18n-foundation | Completed | Spec package and handoffs created. |
+| slice-01-ai-run-status-resume | Completed | Localized current run/status/resume wrappers; full watcher runtime remains in v36. |
+| slice-02-ai-agent-models | Completed | Localized agent/model wrappers and selectors while preserving profile/JSON contracts. |
+| slice-03-ai-planner-approval-review | Completed | Localized planner/review/approval wrappers while preserving prompt and artifact contracts. |
+| slice-04-ai-execution-pr | Completed | Localized execution and PR wrappers while preserving git/gh semantics. |
+| slice-05-ai-tests-smokes | Completed | Full tests, package smoke, spec validation, and evidence completed. |
+| slice-06-ai-prepare-context-progress-i18n-fix | Completed | Fixed mixed-language live progress in onboard and planner prepare-context paths. |
+
+## Current Blockers
+
+- No blockers for v41. `ai run watch` runtime remains a separate planned dependency in `specs/quiver-v36-ai-run-watch-portable`.

package/specs/quiver-v41-cli-i18n-ai-lifecycle/pr.md

@@ -0,0 +1,103 @@
+# PR - Quiver v41 CLI i18n AI Lifecycle
+
+## Title
+
+QUIVER-41: CLI i18n for AI lifecycle commands
+
+## Summary
+
+- Localizes AI lifecycle command wrappers in English and Spanish across run state, agent/model configuration, planner approvals, execution, and PR flows.
+- Preserves provider prompts, prompt-only output, JSON/JSONL surfaces, run ids, provider/model ids, command snippets, git/gh behavior, artifacts, and secret redaction.
+- Closes the v41 spec with full test, package smoke, spec validation, and evidence.
+
+## Scope
+
+- Included: `ai run create`, `ai run close`, `ai status`, `ai resume`, `ai approvals`, `ai agent`, `ai models`, `ai plan`, `ai revise`, `ai review-plan`, `ai repair-plan`, `ai approve`, `ai execute-slice`, `ai execute-plan`, and `ai pr`.
+- Included: human output, progress checks, selector prompts, dry-run reports, actionable errors, review wrappers, and no-TTY guidance.
+- Excluded: provider-generated content translation, provider adapter changes, credential behavior changes, and full `ai run watch` runtime implementation owned by v36.
+
+## Files
+
+- `src/create-quiver/index.js`
+- `src/create-quiver/commands/ai.js`
+- `src/create-quiver/lib/ai/execution-plan.js`
+- `src/create-quiver/lib/ai/github.js`
+- `src/create-quiver/lib/i18n/messages/en.js`
+- `src/create-quiver/lib/i18n/messages/es.js`
+- `tests/commands/ai-*.test.js`
+- `tests/lib/i18n-catalog.test.js`
+- `specs/quiver-v41-cli-i18n-ai-lifecycle/**`
+
+## How to Test (DETAILED - REQUIRED)
+
+### Required Environment
+
+- Node.js and npm supported by this repository.
+- Local checkout of this branch.
+- No live AI provider credentials are required for the automated tests; provider-backed flows use dry-run, print-prompt, or injected provider runners.
+- `gh` is only needed for PR preflight/create flows, not for the full unit suite.
+
+### Worktree Access
+
+```bash
+cd /Users/fabrijk/Documents/Work/Proyectos\ Personales/nika/frameworks/quiver-v41
+git status --short --branch
+```
+
+Expected: branch `feature/QUIVER-41-01-v41-ai-run-status-resume` with no uncommitted changes.
+
+### Run the Project
+
+```bash
+node bin/create-quiver.js --help
+node bin/create-quiver.js --version
+node bin/create-quiver.js version --lang es
+```
+
+Expected: CLI starts normally; Spanish human version output is localized while semver surfaces remain stable.
+
+### Use Cases
+
+```bash
+node bin/create-quiver.js ai models list --lang es
+node bin/create-quiver.js ai agent doctor --lang es --dry-run
+node bin/create-quiver.js ai plan --phase acceptance --input requirements.md --dry-run --lang es
+node bin/create-quiver.js ai execute-plan --dry-run --lang es
+node bin/create-quiver.js ai pr --dry-run --lang es --ssh-host-alias github-personal --identity-file ~/ssh/github-personal --input specs/quiver-v41-cli-i18n-ai-lifecycle/pr.md
+```
+
+Expected: human wrappers are localized; commands, paths, provider ids, model ids, and machine-readable outputs remain exact. The `requirements.md` command requires an existing input file if run outside the test fixtures.
+
+### Technical Verification
+
+```bash
+node --test tests/**/*.test.js
+npm run package:quiver
+npm run smoke:create-quiver
+node bin/create-quiver.js spec validate specs/quiver-v41-cli-i18n-ai-lifecycle --strict
+node bin/create-quiver.js check-slice specs/quiver-v41-cli-i18n-ai-lifecycle/slices/slice-05-ai-tests-smokes/slice.json --local
+git diff --check
+```
+
+Expected: all commands pass.
+
+## Evidence
+
+- PASS `node --test tests/**/*.test.js` (`599/599`)
+- PASS `npm run package:quiver` (`create-quiver-0.15.3.tgz`)
+- PASS `npm run smoke:create-quiver`
+- PASS `node bin/create-quiver.js spec validate specs/quiver-v41-cli-i18n-ai-lifecycle --strict`
+- PASS `node bin/create-quiver.js check-slice specs/quiver-v41-cli-i18n-ai-lifecycle/slices/slice-05-ai-tests-smokes/slice.json --local`
+- PASS `node bin/create-quiver.js check-slice specs/quiver-v41-cli-i18n-ai-lifecycle/slices/slice-06-ai-prepare-context-progress-i18n-fix/slice.json --local`
+- PASS `git diff --check`
+
+## Rollback
+
+- Revert this PR to restore the previous AI lifecycle human output behavior.
+- No migrations, external service changes, provider credential changes, or package manager changes are included.
+
+## Risks / Notes
+
+- Blast radius is the human CLI UX around AI lifecycle commands; automation contracts are covered by JSON/JSONL and prompt stability tests.
+- Full `ai run watch` runtime remains outside v41 and is tracked by `specs/quiver-v36-ai-run-watch-portable`.
+- The final validation found and fixed mixed-language progress in `ai onboard` / `ai prepare-context --with-planner` under `QUIVER-41-06`.

package/specs/quiver-v41-cli-i18n-ai-lifecycle/slices/slice-00-ai-i18n-foundation/CLOSURE_BRIEF.md

@@ -0,0 +1,10 @@
+# CLOSURE_BRIEF - slice-00 AI i18n foundation
+
+## Summary
+
+Documentation package created. No runtime implementation was performed.
+
+## Validation
+
+- Pending `npx create-quiver spec validate specs/quiver-v41-cli-i18n-ai-lifecycle --strict`.
+- Pending `git diff --check`.

package/specs/quiver-v41-cli-i18n-ai-lifecycle/slices/slice-00-ai-i18n-foundation/EXECUTION_BRIEF.md

@@ -0,0 +1,21 @@
+# EXECUTION_BRIEF - slice-00 AI i18n foundation
+
+## Context
+
+AI lifecycle commands need their own migration plan because provider prompts and artifacts must remain stable.
+
+## Objective
+
+Create the v41 AI lifecycle i18n spec package and handoffs.
+
+## Acceptance Criteria
+
+- v41 root documents exist.
+- Each planned slice has execution and closure briefs.
+- Runtime code remains untouched.
+
+## Completion Checklist
+
+- [x] Spec package created.
+- [x] Handoffs created.
+- [x] Runtime implementation deferred.

package/specs/quiver-v41-cli-i18n-ai-lifecycle/slices/slice-00-ai-i18n-foundation/slice.json

@@ -0,0 +1,53 @@
+{
+  "slice_id": "slice-00-ai-i18n-foundation",
+  "ticket": "QUIVER-41-00",
+  "type": "documentation",
+  "title": "AI i18n foundation",
+  "objective": "Create the approved v41 AI lifecycle i18n spec package and handoffs.",
+  "description": "Documents migration of AI lifecycle command wrappers to en and es.",
+  "git": {
+    "branch_type": "docs",
+    "base_branch": "main",
+    "branch_slug": "v41-ai-i18n-foundation",
+    "branch_name": "docs/QUIVER-41-00-v41-ai-i18n-foundation"
+  },
+  "files": [
+    "specs/quiver-v41-cli-i18n-ai-lifecycle/**"
+  ],
+  "expected_read_paths": [
+    "docs/INDEX.md",
+    "docs/CLI_UX_GUIDE.md",
+    "docs/reference/commands.md",
+    "specs/quiver-v37-cli-i18n-foundation/SPEC.md"
+  ],
+  "allowed_write_paths": [
+    "specs/quiver-v41-cli-i18n-ai-lifecycle/**"
+  ],
+  "depends_on": [],
+  "parallel_safe": "no",
+  "parallel_safe_reason": "This creates the roadmap for v41.",
+  "must": [
+    "Create v41 package.",
+    "Document provider prompt guardrails.",
+    "Keep runtime code untouched."
+  ],
+  "not_included": [
+    "Runtime command migration."
+  ],
+  "acceptance": [
+    "Spec package exists.",
+    "Slice handoffs exist.",
+    "Runtime code is untouched."
+  ],
+  "tests": [
+    "npx create-quiver spec validate specs/quiver-v41-cli-i18n-ai-lifecycle --strict",
+    "git diff --check"
+  ],
+  "validation_hints": [
+    "Strict validation requires completed slices to appear in EVIDENCE_REPORT.md."
+  ],
+  "estimated_hours": 1,
+  "status": "completed",
+  "completed_at": "2026-05-28T00:00:00.000Z",
+  "blocked_reason": null
+}

package/specs/quiver-v41-cli-i18n-ai-lifecycle/slices/slice-01-ai-run-status-resume/CLOSURE_BRIEF.md

@@ -0,0 +1,13 @@
+# CLOSURE_BRIEF - slice-01 AI run status and resume
+
+## Summary
+
+Completed. `ai run create`, `ai run close`, `ai status`, `ai resume`, and `ai approvals` now route through the shared language option for human wrapper output. Run ids, phases, statuses, commands, paths, and approval candidate versions remain exact.
+
+`ai run watch` is not implemented in the current runtime; the current unsupported-subcommand wrapper is localized in `en` and `es`. Full watcher runtime and Ctrl+C semantics remain owned by `specs/quiver-v36-ai-run-watch-portable`.
+
+## Validation
+
+- [x] `node --test tests/commands/ai-run-state.test.js`
+- [x] `node --test tests/lib/i18n-catalog.test.js`
+- [x] `git diff --check`

package/specs/quiver-v41-cli-i18n-ai-lifecycle/slices/slice-01-ai-run-status-resume/EXECUTION_BRIEF.md

@@ -0,0 +1,22 @@
+# EXECUTION_BRIEF - slice-01 AI run status and resume
+
+## Context
+
+Run status and watcher output expose AI progress and must remain clear in both languages.
+
+## Objective
+
+Localize `ai run create`, `ai run watch`, `ai status`, and `ai resume` wrapper output.
+
+## Acceptance Criteria
+
+- Human output supports `en` and `es`.
+- Run ids, paths, and status codes remain stable.
+- JSON and JSONL modes remain pure and parseable.
+- Watcher cancellation semantics remain unchanged.
+
+## Completion Checklist
+
+- [ ] Run/status messages cataloged.
+- [ ] JSON/JSONL regression tests preserved.
+- [ ] Watch tests cover both languages.

package/specs/quiver-v41-cli-i18n-ai-lifecycle/slices/slice-01-ai-run-status-resume/slice.json

@@ -0,0 +1,68 @@
+{
+  "slice_id": "slice-01-ai-run-status-resume",
+  "ticket": "QUIVER-41-01",
+  "type": "feature",
+  "title": "AI run status and resume",
+  "objective": "Localize AI run status and watcher wrapper output while preserving JSON and JSONL contracts.",
+  "description": "Covers run creation/status/resume/watch human output.",
+  "git": {
+    "branch_type": "feature",
+    "base_branch": "main",
+    "branch_slug": "v41-ai-run-status-resume",
+    "branch_name": "feature/QUIVER-41-01-v41-ai-run-status-resume"
+  },
+  "files": [
+    "src/create-quiver/index.js",
+    "src/create-quiver/commands/ai.js",
+    "src/create-quiver/lib/ai/run-state.js",
+    "src/create-quiver/lib/i18n/messages/en.js",
+    "src/create-quiver/lib/i18n/messages/es.js",
+    "tests/commands/ai-run-state.test.js"
+  ],
+  "expected_read_paths": [
+    "src/create-quiver/commands/ai.js",
+    "src/create-quiver/lib/ai/run-state.js",
+    "src/create-quiver/lib/i18n/**",
+    "specs/quiver-v41-cli-i18n-ai-lifecycle/SPEC.md"
+  ],
+  "allowed_write_paths": [
+    "src/create-quiver/commands/ai.js",
+    "src/create-quiver/index.js",
+    "src/create-quiver/lib/ai/**",
+    "src/create-quiver/lib/i18n/**",
+    "tests/commands/ai-run-state.test.js",
+    "tests/commands/ai-run-watch.test.js",
+    "specs/quiver-v41-cli-i18n-ai-lifecycle/**"
+  ],
+  "depends_on": [
+    "slice-00-ai-i18n-foundation"
+  ],
+  "parallel_safe": "no",
+  "parallel_safe_reason": "Establishes AI lifecycle localization patterns for v41.",
+  "must": [
+    "Localize human run/status outputs.",
+    "Keep run ids and status codes stable.",
+    "Keep JSON/JSONL pure."
+  ],
+  "not_included": [
+    "Planner and provider prompt localization."
+  ],
+  "acceptance": [
+    "Run/status/resume outputs render in en and es.",
+    "The current unsupported watch wrapper error renders in en and es until the v36 watcher runtime lands.",
+    "JSON and JSONL tests remain stable.",
+    "Watcher Ctrl+C behavior is unchanged because v36 owns watcher runtime implementation."
+  ],
+  "tests": [
+    "node --test tests/commands/ai-run-state.test.js",
+    "node --test tests/lib/i18n-catalog.test.js",
+    "git diff --check"
+  ],
+  "validation_hints": [
+    "If ai-run-watch tests do not exist yet, add focused tests with temporary run fixtures."
+  ],
+  "estimated_hours": 7,
+  "status": "completed",
+  "blocked_reason": "",
+  "completed_at": "2026-05-29T17:29:09Z"
+}

package/specs/quiver-v41-cli-i18n-ai-lifecycle/slices/slice-02-ai-agent-models/CLOSURE_BRIEF.md

@@ -0,0 +1,12 @@
+# CLOSURE_BRIEF - slice-02 AI agents and models
+
+## Summary
+
+Completed. `ai agent set/list/show/doctor/repair` and `ai models list` now localize human wrapper output in `en` and `es`. Provider ids, model ids, stored profile JSON, and `--json` output remain stable.
+
+## Validation
+
+- [x] `node --test tests/commands/ai-agent.test.js`
+- [x] `node --test tests/commands/ai-models.test.js`
+- [x] `node --test tests/lib/i18n-catalog.test.js`
+- [x] `git diff --check`

package/specs/quiver-v41-cli-i18n-ai-lifecycle/slices/slice-02-ai-agent-models/EXECUTION_BRIEF.md

@@ -0,0 +1,22 @@
+# EXECUTION_BRIEF - slice-02 AI agents and models
+
+## Context
+
+Agent/model commands mix human labels with technical provider/model identifiers.
+
+## Objective
+
+Localize `ai agent set`, `ai agent doctor`, `ai agent repair`, and `ai models list`.
+
+## Acceptance Criteria
+
+- Human labels render in `en` and `es`.
+- Provider names, model ids, and custom model ids remain exact.
+- `--json` output remains stable.
+- Interactive selectors localize prompts and choices without changing stored values.
+
+## Completion Checklist
+
+- [ ] Agent/model messages cataloged.
+- [ ] Selector prompts tested.
+- [ ] Stored profile regression tests preserved.

package/specs/quiver-v41-cli-i18n-ai-lifecycle/slices/slice-02-ai-agent-models/slice.json

@@ -0,0 +1,66 @@
+{
+  "slice_id": "slice-02-ai-agent-models",
+  "ticket": "QUIVER-41-02",
+  "type": "feature",
+  "title": "AI agents and models",
+  "objective": "Localize agent/model command UX while preserving technical identifiers and stored profile contracts.",
+  "description": "Covers AI agent setup, doctor, repair, and model catalog commands.",
+  "git": {
+    "branch_type": "feature",
+    "base_branch": "main",
+    "branch_slug": "v41-ai-agent-models",
+    "branch_name": "feature/QUIVER-41-02-v41-ai-agent-models"
+  },
+  "files": [
+    "src/create-quiver/index.js",
+    "src/create-quiver/commands/ai.js",
+    "src/create-quiver/lib/ai/model-catalog.js",
+    "src/create-quiver/lib/i18n/messages/en.js",
+    "src/create-quiver/lib/i18n/messages/es.js",
+    "tests/commands/ai-agent.test.js",
+    "tests/commands/ai-models.test.js"
+  ],
+  "expected_read_paths": [
+    "src/create-quiver/commands/ai.js",
+    "src/create-quiver/lib/ai/models.js",
+    "src/create-quiver/lib/i18n/**"
+  ],
+  "allowed_write_paths": [
+    "src/create-quiver/index.js",
+    "src/create-quiver/commands/ai.js",
+    "src/create-quiver/lib/ai/**",
+    "src/create-quiver/lib/i18n/**",
+    "tests/commands/ai-agent.test.js",
+    "tests/commands/ai-models.test.js",
+    "specs/quiver-v41-cli-i18n-ai-lifecycle/**"
+  ],
+  "depends_on": [
+    "slice-01-ai-run-status-resume"
+  ],
+  "parallel_safe": "conditional",
+  "parallel_safe_reason": "Can overlap with planner work if command sections are isolated.",
+  "must": [
+    "Localize prompts and human labels.",
+    "Keep stored model/provider values exact.",
+    "Keep JSON stable."
+  ],
+  "not_included": [
+    "Provider execution behavior changes."
+  ],
+  "acceptance": [
+    "Agent/model commands render in en and es.",
+    "Profile JSON is unchanged by language.",
+    "Interactive selectors store technical ids."
+  ],
+  "tests": [
+    "node --test tests/commands/ai-agent.test.js tests/commands/ai-models.test.js tests/lib/i18n-catalog.test.js",
+    "git diff --check"
+  ],
+  "validation_hints": [
+    "If model list tests are embedded in cli-contract tests, update the nearest owning test."
+  ],
+  "estimated_hours": 7,
+  "status": "completed",
+  "blocked_reason": "",
+  "completed_at": "2026-05-29T17:36:48Z"
+}

package/specs/quiver-v41-cli-i18n-ai-lifecycle/slices/slice-03-ai-planner-approval-review/CLOSURE_BRIEF.md

@@ -0,0 +1,12 @@
+# CLOSURE_BRIEF - slice-03 AI planner, approvals, and review
+
+## Summary
+
+Completed. `ai plan`, `ai revise`, `ai review-plan`, `ai repair-plan`, `ai approve`, and approval status wrappers now receive the effective language and localize human dry-run, prompt-only, approval, review, repair, progress, and actionable wrapper messages. Provider prompt bodies, phase names, version ids, commands, draft paths, and approval artifacts remain exact.
+
+## Validation
+
+- [x] `node --test tests/commands/ai-plan-spec-phase.test.js`
+- [x] `node --test tests/commands/flow.test.js`
+- [x] `node --test tests/lib/i18n-catalog.test.js`
+- [x] `git diff --check`

package/specs/quiver-v41-cli-i18n-ai-lifecycle/slices/slice-03-ai-planner-approval-review/EXECUTION_BRIEF.md

@@ -0,0 +1,23 @@
+# EXECUTION_BRIEF - slice-03 AI planner, approvals, and review
+
+## Context
+
+Planner commands are provider-backed and use review/approval gates. Quiver wrapper messages can be localized, but provider prompts and generated drafts remain stable unless explicitly changed.
+
+## Objective
+
+Localize `ai plan`, `ai revise`, `ai review-plan`, `ai repair-plan`, `ai approve`, and approval status surfaces.
+
+## Acceptance Criteria
+
+- Human wrapper output supports `en` and `es`.
+- Provider prompt text printed by `--print-prompt` remains exact.
+- Approval candidates, review summaries, and blocking messages are localized.
+- Draft paths, phase names, version ids, and commands remain exact.
+- JSON/status artifacts remain stable.
+
+## Completion Checklist
+
+- [ ] Planner wrapper strings cataloged.
+- [ ] Approval/review strings cataloged.
+- [ ] `--print-prompt` regression tests preserved.