Actions

GPTClaudeDeepSeek··▲ 1,373 copies·updated 2026-07-14

actions.prompt

# 99 — Human Manual Action Items
# Things only CS can do (IDE setup, external accounts, app stores, etc.)
# Any agent session can append items here.
# PM reviews and organizes by priority each cycle.

---

## PENDING — Work through by priority

### P0 — Ship & Distribute
- [x] **Demo GIF (#131)** — DONE: `scripts/demo/run-demo.sh` + `scripts/demo/record-demo.sh` created. Run demo, record with asciinema, convert to GIF with `agg`.
- [ ] **Record actual GIF** — Run `bash scripts/demo/record-demo.sh`, convert to GIF, add to README
- [ ] **Run benchmark dry-run** — `bash scripts/benchmark/run-orchestration-bench.sh --project-dir . --dry-run`
- [ ] **Tag v0.2.0** — CLAUDE.md says v0.2.0 is fully wired (9/9 modules, 278 tests, CTO 8/8, QA 8/8, Security 6/6). Tag it.

### P1 — Incomplete Features
- [x] **#169 Benchmark completion** — DONE: `run-orchestration-bench.sh` (standalone orchestration benchmark) + `run-comparison.sh` (compare two runs). Original SWE-bench runner also exists.
- [x] **#121 Docs site foundation** — DONE: `docs-site/` with mint.json + 6 MDX pages (intro, quickstart, 3 concepts, 1 API ref)
- [ ] **Connect Mintlify to GitHub** — needed for docs auto-deploy. Go to mintlify.com/start, point to `docs-site/`
- [ ] **Logo/branding assets** — web agent needs logo SVGs in `docs-site/logo/`
- [ ] **BUG-026 (#190)** — `prev` uninitialized in agent-health-report.sh:48. Fix: add `prev=""` on line 44.
- [ ] **Fix pre-existing test failures** — test-e2e-dry-run.sh (output format mismatch), test-prompt-template.sh (partial)

### P2 — Future Product
- [ ] Tauri desktop app (#97-101)
- [ ] Pixel Agents integration (#94-96)
- [ ] macOS compatibility (#143)
- [ ] Model fallback routing (#160)
- [ ] Max cycles env override (#159)

---

## DONE

- [x] **2026-04-03** 5-cycle orchestration sprint: benchmark suite (run-orchestration-bench.sh + run-comparison.sh), Mintlify docs site (7 files), demo script (run-demo.sh + record-demo.sh), 42 extended v0.2.0 tests, PM update.
- [x] **2026-04-02** Verified all P0.5 wiring items complete in auto-agent.sh. Updated actions file to reflect reality.
- [x] **2026-03-31** v0.2.0 modules FULLY WIRED into auto-agent.sh: timeout (#88), dry-run (#90), review-phase (#145), model tiering (#147), worktree (#168), session tracker (BUG-025), conditional activation, differential context, quality gates. All use guarded `type -t` checks.
- [x] **2026-03-31** STOP ORCHESTRATOR resolved — CTO review queue cleared, v0.2.0 modules wired, orchestrator unblocked.
- [x] **2026-03-31** Closed 20 completed + 4 obsolete issues. Fixed 4 bugs (BUG-024, #151, #156, #158). Reopened 7 genuinely incomplete issues.
- [x] **2026-03-31** BUG-025 fixed — `orch_tracker_*` renamed to `orch_session_*` in auto-agent.sh (zero `orch_tracker_` references remain).
- [x] **2026-03-30** v0.1.0 TAGGED (6a02462) and PUSHED to origin (confirmed on remote).
- [x] **2026-03-30** All 22 modules wired (sourced). 5 efficiency scripts wired. 23/23 tests pass.
- [x] **2026-03-30** 10 bugs fixed: BUG-013, QA-F001, SS-01, CS-01, LINT-01-04, BUG-012. Shared resources synced (22 modules).
- [x] **2026-03-30** CLAUDE.md rewritten, .gitignore cleaned, agents.conf symlinked, orphans removed.
- [x] **2026-03-30** ~12 agent cycles ran: single-agent.sh, benchmark framework, qmd-refresher, prompt-template, task-decomposer, init-project built.
- [x] **2026-03-29** v0.1.0 all blockers resolved: security FULL PASS, QA PASS, CTO 8/8.

when to use it

Community prompt sourced from the open-source GitHub repo ChetanSarda99/openOrchyStraw (MIT). A "Actions" style prompt — adapt the placeholders and specifics to your task. Imported as-is and not independently retested here, so check the output before relying on it.