project-review
Skill43 starsChanged 7 months agoMITView on GitHub
Strategic project review — auto-discovers strategy docs (vision, roadmap, decisions, notes), produces five-dimension snapshot. Use for direction review, alignment check, milestone status, pre-meeting overview. Triggers: 'project review', 'strategic review', 'where is the project at', '项目全景', '审视战略', '项目现在什么状态'. Not for code/PR review.
17 other projects use a version of this.
Adopted by
Most recently taken up by
- loupe-ac/loupetheir own version, first seen Nov 16, 2026
- lwmxiaobei/xiaobei-skillstheir own version, first seen Nov 15, 2026
- Eleven-M/flowmindtheir own version, first seen Nov 8, 2026
- ubermuda/loupetheir own version, first seen Nov 8, 2026
- 70asunflower/70asunflower-skillstheir own version, first seen Nov 8, 2026
- SWEStash/swe-workflow-skillstheir own version, first seen Nov 8, 2026
Other versions17 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 17
- jerry-ai-dev/MODULAR-RAG-MCP-SERVER★ 1.2ksame name
- lwmxiaobei/xiaobei-skills★ 156same name
- 5throck/ai-workspace-standards★ 25same name
- rittmananalytics/wire-plugin★ 14same name
- fqmyysjjd/agent-feed★ 4same name
- ila/duckdb-claude-skills★ 4same name
- loupe-ac/loupe★ 2same name
- ubermuda/loupe★ 2same name
- 70asunflower/70asunflower-skills★ 1same name
- davidsolheim/skills★ 1same name
7 more not listed
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository17 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- bibtex-fetchSkill
- context-auditSkill
- daily-summarySkill
- experiment-registrySkill
- handoffSkill
- init-projectSkill
- notion-lifeosSkill
- paper-analyzerSkill
- paper-self-reviewSkill
- remote-cluster-agentSkill
- research-leadSkill
- session-teamSkill
- sync-docsSkill
- team-digestSkill
- web-fetcherSkill
The file
We split this skill into the rules it gives an agent
7 rules. 0 are common, 4 are distinctive, and 3 are very specific to this project.
0
Common
Most of the 12 edited versions we read keep these.
4
Distinctive
This author's own choices. Worth a look.
3
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 16
If it exists, read each path listed — the project maintainer has already specified which documents constitute the "strategic picture", no …
- Distinctive: line 18
If it doesn't exist, fall through to Phase 1b (auto-discovery).
- Very specific: line 50
Search the following paths, excluding third_party/, node_modules/, .venv/, vendor/:
- Very specific: line 63
Use exact paths with sequential fallback — avoid / recursive globs, which tend to match files under third_party/ and produce noise.
- Distinctive: line 65
If none of the above paths exist, prompt the user:
- Distinctive: line 75
If no vision document exists, infer project goals from README (mark as inferred)
- Very specific: line 125
If documents were found via Phase 1b auto-discovery (rather than the config file), append a recommended configuration at the end of the …
“Common” is measured: kept by at least 6 in 10 of the 12 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Project Review
Reads the project's strategy document ecosystem and generates a panoramic snapshot. Useful before team meetings, when feeling lost about direction, or for periodic retrospectives — helps you quickly understand where the project is and where it should go.
Execution Skeleton
Phase 1: Load Document Manifest
First check if the project has an explicit configuration: read docs/strategy/.review-sources.md.
If it exists, read each path listed — the project maintainer has already specified which documents constitute the "strategic picture", no guessing needed.
If it doesn't exist, fall through to Phase 1b (auto-discovery).
Configuration Format (docs/strategy/.review-sources.md)
# Project Review Sources
## Core
- docs/strategy/vision.md
- docs/strategy/roadmap.md
- docs/strategy/paper-outline.md
## Decisions
- docs/strategy/decisions/log.md
## Related Work
- docs/strategy/related-work/_index.md
## Meetings (latest 2)
- docs/strategy/meetings/
## Experiments
- .claude/knowledge/experiments.md
## Progress
- HANDOFF.mdEach entry is a path relative to project root. Directory paths (ending with /) mean "read the 2 most recently modified files in that directory". Lines starting with # are section headers for grouping.
Phase 1b: Auto-Discovery (fallback when no config exists)
Search the following paths, excluding third_party/, node_modules/, .venv/, vendor/:
| Document Type | Search Paths (by priority) | Purpose |
| Vision | docs/strategy/vision.md → docs/vision.md → VISION.md | Core positioning and assumptions |
| Roadmap | docs/strategy/roadmap.md → docs/roadmap.md → ROADMAP.md | Milestones and timeline |
| Paper Outline | docs/strategy/paper-outline.md | Paper outline (research projects) |
| Related Work | _index.md or all .md under docs/strategy/related-work/ | Related work coverage |
| Decisions | docs/strategy/decisions/log.md → docs/strategy/decisions/ → docs/adr/ | Decision records |
| Meetings | 2 most recent .md under docs/strategy/meetings/ (exclude _template.md) | Meeting notes |
| Experiments | .claude/knowledge/experiments.md | Experiment log |
| Progress | HANDOFF.md (project root) | Current progress |
Use exact paths with sequential fallback — avoid **/ recursive globs, which tend to match files under third_party/ and produce noise.
If none of the above paths exist, prompt the user:
"No strategy documents found. Consider creating
docs/strategy/.review-sources.mdto specify document locations, or createvision.mdandroadmap.mdunderdocs/strategy/."
Phase 2: Five-Dimension Analysis
Generate analysis based on the discovered documents. Only produce a dimension when supporting documents exist — skip dimensions without data, because speculation is more dangerous than silence.
1. Vision Check
- Is current work aligned with core objectives / assumptions?
- Which assumptions are validated, challenged, or untested?
- If no vision document exists, infer project goals from README (mark as inferred)
2. Roadmap Status
- Progress overview (done / in-progress / planned)
- Current phase's position in the overall roadmap
- Any work that deviates from the plan
3. Bottleneck Identification
- Extract current blockers from experiment results and decision records
- Distinguish technical bottlenecks (e.g., unstable API) from research/business bottlenecks (e.g., invalidated assumptions)
4. Related Work Gap (research projects)
- Which topics lack related work coverage?
- Are there newly published works that need attention?
5. Next Steps
- Recommendations based on the above analysis
- Priority ordering: urgent vs. important
Phase 3: Output
========================================
PROJECT REVIEW SNAPSHOT
<project-name> | YYYY-MM-DD
========================================
## 1. Vision Check
...
## 2. Roadmap Status
| ID | Milestone | Status | Notes |
...
## 3. Bottleneck Identification
...
## 4. Related Work Gap
...
## 5. Next Steps
1. ...
2. ...
3. ...
========================================First-Use Onboarding
If documents were found via Phase 1b auto-discovery (rather than the config file), append a recommended configuration at the end of the output so the user can quickly create .review-sources.md:
Tip: Create docs/strategy/.review-sources.md to precisely specify document
locations and avoid false matches. Recommended config:
<generate config content based on actually discovered files>Constraints
- Read-only — do not modify any documents. This skill's value is providing perspective, not making decisions for the user.
- Output to terminal — do not create files unless the user explicitly asks to save. Avoid "documents writing documents" noise.
- Follow the user's language — respond in whichever language the user uses.
- Skip dimensions that lack supporting documents — silence is more valuable than false confidence.
- This skill does not replace weekly-report (for reporting) or meeting-slides (for presentations) — it's for personal/team strategic reflection.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

