review-plan

Name: review-plan
Author: tobihagemann

4installs29forks

How to install review-plan?

tobihagemann/turbo/review-plan

$npx -y skills add tobihagemann/turbo --skill review-plan

Installs into the current project.

Summary

Review a planning artifact (plan, shells, or spec) by running internal and peer reviews in parallel and returning combined findings. Use when the user asks to \"review my plan\", \"review my shells\", \"review my spec\", \"check my plan\", \"check my shells\", \"check my spec\",

Files · 1

View on GitHub

SKILL.md

1	# Review Plan
2
3	Review a planning artifact against type-specific criteria. Runs internal review and `/peer-review` in parallel by default. Returns combined structured findings.
4
5	## Step 1: Determine Artifact Type and Resolve
6
7	### Determine Artifact Type
8
9	1. Explicit argument — If the user specified a type (e.g., "review shells", "review spec"), use it. No argument defaults to plan.
10	2. Conversation context — If artifact text or a path is already in context, infer the type.
11	3. Auto-detect — Check `.turbo/` for existing artifacts. If multiple types exist, pick the one with the most recently modified file.
12
13	### Resolve the Artifact
14
15	#### Plan (default)
16
17	1. Plan text in conversation — use it
18	2. Explicit path — read it
19	3. Explicit slug — resolve to `.turbo/plans/<slug>.md`
20	4. Single file — Glob `.turbo/plans/*.md`. If exactly one file exists, use it
21	5. Most recent — most recently modified file
22	6. Legacy fallback — `.turbo/plan.md` if `.turbo/plans/` does not exist
23	7. Nothing found — stop and state that no artifact was found to review
24
25	#### Shells
26
27	1. Shell text in conversation — use it
28	2. Explicit spec slug — Glob `.turbo/shells/<slug>-*.md`
29	3. Explicit spec path — derive slug from filename, glob as above
30	4. Single spec — Glob `.turbo/specs/*.md`. If exactly one, derive slug and glob for shells
31	5. Most recent spec — most recently modified spec, derive slug and glob
32	6. Nothing found — stop and state that no artifact was found to review
33
34	For shells, read each shell file and parse its YAML frontmatter (`spec`, `depends_on`). Read the source spec from the `spec` field.
35
36	#### Spec
37
38	1. Spec text in conversation — use it
39	2. Explicit path — read it
40	3. Explicit slug — resolve to `.turbo/specs/<slug>.md`
41	4. Single file — Glob `.turbo/specs/*.md`. If exactly one, use it
42	5. Most recent — most recently modified
43	6. Legacy fallback — `.turbo/spec.md` if `.turbo/specs/` does not exist
44	7. Nothing found — stop and state that no artifact was found to review
45
46	If multiple candidates exist, pick the most recently modified.
47
48	## Step 2: Run Reviews in Parallel
49
50	Read the reference file for the resolved type:
51
52	- Plan — [references/plan-review.md](references/plan-review.md)
53	- Shells — [references/shells-review.md](references/shells-review.md)
54	- Spec — [references/spec-review.md](references/spec-review.md)
55
56	Skip peer review when the caller asked (e.g., "without peer review", "no peer", "internal only"). For shells, the internal review focuses on structural wiring and skips the project context read.
57
58	Use the Agent tool to launch all agents below in a single assistant message so they run concurrently. Run them in the foreground so all their results return in this turn. Each Agent call uses `model: "opus"`. That is two Agent tool calls when peer review is active (one internal + one peer), or one Agent tool call when peer review is skipped. Every agent's prompt must direct it to treat the shared working tree and its git index as read-only and to assess findings by reading and reasoning. For a check that genuinely requires mutating code (such as testing whether a finding holds), the agent works in an isolated `git worktree` it discards afterward.
59
60	- Internal Agent: Pass the artifact text and the reference file's content; the subagent reads project context (CLAUDE.md, relevant codebase files) before applying criteria, then returns findings in the output format below.
61	- Peer review Agent (unless skipping): Launch an Agent tool call whose prompt instructs the subagent to invoke `/peer-review` via the Skill tool with a request describing: (a) the artifact under review; (b) the criteria live in `~/.claude/skills/review-plan/references/<type>-review.md` for the resolved type from Step 1 — the reviewer should read that file directly and use its priority scale; (c) the Overall Verdict should use the `Readiness: <ready \| needs revision>` label. The prompt must also state explicitly that the subagent's final assistant message must contain the verbatim findings text `/peer-review` produced.
62
63	Aggregate findings with attribution (reviewer: "internal" or "peer"). Present them in the output format below.
64
65	Then use the TaskList tool and proceed to any remaining task.
66
67	## Output Format
68
69	Format each finding as:
70
71	```
72	### [P<N>] <title (imperative, ≤80 chars)>
73
74	Section: <plan section, shell number(s), or spec section>
75	Reviewer: <internal \| peer>
76
77	<one paragraph explaining the issue and its impact>
78	```
79
80	After all findings

People who liked this also viewed

6 picks

Type

review-plan

4installs29forks

View on GitHub

How to install review-plan?

tobihagemann/turbo/review-plan

$npx -y skills add tobihagemann/turbo --skill review-plan

Installs into the current project.

Summary

Files · 1

View on GitHub

SKILL.md

1	# Review Plan
2
3	Review a planning artifact against type-specific criteria. Runs internal review and `/peer-review` in parallel by default. Returns combined structured findings.
4
5	## Step 1: Determine Artifact Type and Resolve
6
7	### Determine Artifact Type
8
9	1. Explicit argument — If the user specified a type (e.g., "review shells", "review spec"), use it. No argument defaults to plan.
10	2. Conversation context — If artifact text or a path is already in context, infer the type.
11	3. Auto-detect — Check `.turbo/` for existing artifacts. If multiple types exist, pick the one with the most recently modified file.
12
13	### Resolve the Artifact
14
15	#### Plan (default)
16
17	1. Plan text in conversation — use it
18	2. Explicit path — read it
19	3. Explicit slug — resolve to `.turbo/plans/<slug>.md`
20	4. Single file — Glob `.turbo/plans/*.md`. If exactly one file exists, use it
21	5. Most recent — most recently modified file
22	6. Legacy fallback — `.turbo/plan.md` if `.turbo/plans/` does not exist
23	7. Nothing found — stop and state that no artifact was found to review
24
25	#### Shells
26
27	1. Shell text in conversation — use it
28	2. Explicit spec slug — Glob `.turbo/shells/<slug>-*.md`
29	3. Explicit spec path — derive slug from filename, glob as above
30	4. Single spec — Glob `.turbo/specs/*.md`. If exactly one, derive slug and glob for shells
31	5. Most recent spec — most recently modified spec, derive slug and glob
32	6. Nothing found — stop and state that no artifact was found to review
33
34	For shells, read each shell file and parse its YAML frontmatter (`spec`, `depends_on`). Read the source spec from the `spec` field.
35
36	#### Spec
37
38	1. Spec text in conversation — use it
39	2. Explicit path — read it
40	3. Explicit slug — resolve to `.turbo/specs/<slug>.md`
41	4. Single file — Glob `.turbo/specs/*.md`. If exactly one, use it
42	5. Most recent — most recently modified
43	6. Legacy fallback — `.turbo/spec.md` if `.turbo/specs/` does not exist
44	7. Nothing found — stop and state that no artifact was found to review
45
46	If multiple candidates exist, pick the most recently modified.
47
48	## Step 2: Run Reviews in Parallel
49
50	Read the reference file for the resolved type:
51
52	- Plan — [references/plan-review.md](references/plan-review.md)
53	- Shells — [references/shells-review.md](references/shells-review.md)
54	- Spec — [references/spec-review.md](references/spec-review.md)
55
56	Skip peer review when the caller asked (e.g., "without peer review", "no peer", "internal only"). For shells, the internal review focuses on structural wiring and skips the project context read.
57
58	Use the Agent tool to launch all agents below in a single assistant message so they run concurrently. Run them in the foreground so all their results return in this turn. Each Agent call uses `model: "opus"`. That is two Agent tool calls when peer review is active (one internal + one peer), or one Agent tool call when peer review is skipped. Every agent's prompt must direct it to treat the shared working tree and its git index as read-only and to assess findings by reading and reasoning. For a check that genuinely requires mutating code (such as testing whether a finding holds), the agent works in an isolated `git worktree` it discards afterward.
59
60	- Internal Agent: Pass the artifact text and the reference file's content; the subagent reads project context (CLAUDE.md, relevant codebase files) before applying criteria, then returns findings in the output format below.
61	- Peer review Agent (unless skipping): Launch an Agent tool call whose prompt instructs the subagent to invoke `/peer-review` via the Skill tool with a request describing: (a) the artifact under review; (b) the criteria live in `~/.claude/skills/review-plan/references/<type>-review.md` for the resolved type from Step 1 — the reviewer should read that file directly and use its priority scale; (c) the Overall Verdict should use the `Readiness: <ready \| needs revision>` label. The prompt must also state explicitly that the subagent's final assistant message must contain the verbatim findings text `/peer-review` produced.
62
63	Aggregate findings with attribution (reviewer: "internal" or "peer"). Present them in the output format below.
64
65	Then use the TaskList tool and proceed to any remaining task.
66
67	## Output Format
68
69	Format each finding as:
70
71	```
72	### [P<N>] <title (imperative, ≤80 chars)>
73
74	Section: <plan section, shell number(s), or spec section>
75	Reviewer: <internal \| peer>
76
77	<one paragraph explaining the issue and its impact>
78	```
79
80	After all findings

People who liked this also viewed

6 picks

Type