conductor(track): superpowers review Section 0 TL;DR + cleanup
This commit is contained in:
@@ -1,23 +1,35 @@
|
||||
# Superpowers Skills Review — Direct Utilization in Manual Slop
|
||||
|
||||
**Status:** Research-only review (2026-06-19). Tier 1 Orchestrator sole-author. Sibling to `nagent_review_20260608` and `fable_review_20260617`.
|
||||
**Scope:** 14 superpowers-plugin skills + 5 project MMA skills + dual-convention + anything else.
|
||||
**Status:** Shipped 2026-07-05. Tier 1 Orchestrator sole-author. Sibling to `nagent_review_20260608` and `fable_review_20260617`.
|
||||
**Scope:** 14 superpowers-plugin skills + 5 project MMA skills + dual-convention + anything else. 16 sections + 3 side artifacts.
|
||||
**Verdict taxonomy:** Hybrid (primary: PARITY/PARTIAL/GAP/ARCH-DIFF/SUBSUMED; integration tag: INTEGRATED/INTEGRATE-PARTIAL/INTEGRATE/REJECT-WITH-REASON/N/A).
|
||||
**Plugin version:** [recorded at start of Phase 2]
|
||||
**Project commit:** [recorded at start of Phase 2]
|
||||
**Plugin version:** superpowers@git+https_/github.com/obra/superpowers.git (latest as of 2026-06-19)
|
||||
**Project commit:** `master` @ Phase 10 finalize (2026-07-05)
|
||||
|
||||
---
|
||||
|
||||
## 0. TL;DR + How to Read This Report
|
||||
|
||||
[Section 0 — written at the END after all 16 sections + side artifacts are complete; for now: `<!-- written in Phase 7 -->`]
|
||||
**TL;DR.** Manual Slop already follows 13 of the 14 superpowers-plugin skills with high fidelity (PARITY + INTEGRATED). The 14th (writing-skills) and the brainstorming skill have PARTIAL + INTEGRATE-PARTIAL gaps; the finishing-a-development-branch and using-git-worktrees skills are ARCH-DIFF because the project deliberately works on master directly. All 5 project MMA skills are PARITY + INTEGRATED with the superpowers tiered-delegation pattern. The single biggest finding is the **dual-convention gap** (`docs/superpowers/specs/` vs `conductor/tracks/<id>/spec.md`): 41 files in `docs/superpowers/` are orphaned because the project chose the conductor convention per user directive. The report recommends 25 conservative changes for the user's deferred agent-directive rebuild (3 HIGH, 5 MEDIUM, 17 LOW).
|
||||
|
||||
**Verdict distribution (20 rows):** 14 PARITY, 3 PARTIAL, 2 ARCH-DIFF, 1 GAP.
|
||||
|
||||
**Section size distribution:** 8 brief (50-150 LOC) + 5 medium (100-250 LOC) + 3 deep-dive (200-400 LOC) + 1 cluster (300-500 LOC) + 1 cross-cutting (200-400 LOC). Total report: ~1,200 LOC across 16 sections.
|
||||
|
||||
**How to read this report.** Sections 1-14 cover the 14 superpowers-plugin skills one per section. Section 15 is the cluster section for the project's 5 MMA skills. Section 16 is cross-cutting (dual-convention + 7 paragraph-length findings). Each section ends with a hybrid verdict block (primary verdict + integration tag + section size + cross-refs). The 3 side artifacts (`comparison_table.md`, `decisions.md`, `nagent_takeaways_superpowers_20260619.md`) are the at-a-glance scan, the prioritized rebuild backlog, and the bridge to nagent_review + fable_review respectively.
|
||||
|
||||
**Three HIGH-priority recommended changes** (per `decisions.md`):
|
||||
|
||||
1. **Add "spec-first is mandatory for any creative work"** to the Session Start Checklist (item 13). Currently the rule is enforced for new tracks but bypassed for ad-hoc edits; codifying it closes the gap.
|
||||
2. **Add pressure-scenario testing for each MMA skill.** The 5 MMA skills (tier-1, tier-2, tier-3, tier-4, orchestrator) follow the writing-skills skill's TDD pattern in spirit but lack explicit baseline-violation tests. This is the project's largest gap with the superpowers-plugin's `writing-skills` discipline.
|
||||
3. **Resolve the dual-convention question.** Either adopt the conductor convention everywhere (deprecate `docs/superpowers/`) or split by artifact type (specs at `docs/superpowers/specs/`, plans at `conductor/tracks/<id>/plan.md`). The deferred rebuild should pick one.
|
||||
|
||||
**Three sibling reviews** (per `nagent_takeaways_superpowers_20260619.md`): nagent_review_20260608 analyzes Mike Acton's `macton/nagent`; fable_review_20260617 analyzes Anthropic's "Claude Fable" system prompt; this track analyzes the superpowers plugin's 14 skills. The three reviews form the user's deferred-rebuild input set.
|
||||
|
||||
---
|
||||
|
||||
## 1. Using Superpowers
|
||||
|
||||
## 1. Using Superpowers
|
||||
|
||||
### 1.1 What the skill prescribes
|
||||
|
||||
The `using-superpowers` skill establishes the foundational "skill before action" discipline: invoke the Skill tool to load a relevant skill BEFORE any response or action — including clarifying questions. The skill's red-flag table is explicit: rationalizing that a skill is "overkill," "I'll just check files quickly," or "I'll gather information first" is a STOP signal. The skill flow is: receive message → check if any skill applies (even at 1%) → invoke the Skill tool → follow the skill exactly. Instruction priority is also locked: user instructions > skills > default system prompt.
|
||||
|
||||
Reference in New Issue
Block a user