agentleFS
aka-kikaaka-kika/akakika-skills/skills/project-intelligence/project-worth-my-time/SKILL.md

project-worth-my-time

Skill11 starsChanged 3 months agoMITView on GitHub

Use after PROJECT_GUIDE.md exists, when the user asks if a project is worth their time, should they build vs buy, or invokes project-worth-my-time. Follow-up to explain-new-project—weighs project context (files, git history, chat) to recommend whether to invest in a repo, cherry-pick a feature, park the idea, pass, or use an existing polished product instead. Optionally enriched by a personal-memory MCP (e.g. Pieces) when one is available.

The file

Show its rules and text

We split this skill into the rules it gives an agent

16 rules. About 0 are common, 11 are distinctive, and 5 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈11

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 5

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    If PROJECT_GUIDE.md exists at the project root, read it first (do not re-explore from scratch unless it is stale or missing quick-start …

    line 15
  2. Distinctive:

    If it does not exist, either run the same exploration as explain-new-project or tell the user: *Run the explain-new-project skill first for …

    line 16
  3. Very specific:

    Read project evidence: PROJECT_GUIDE.md, README, manifest, rough maturity signals (last commit, TODOs, half-built features).

    line 24
  4. Distinctive:

    Write the verdict using the template below.

    line 26
  5. Distinctive:

    Use for tone, role, and stated preferences.

    line 40
  6. Very specific:

    Use ask_pieces_ltm with question, topics, open_files (current workspace paths), and related_questions when helpful.

    line 44
  7. Distinctive:

    When the idea is exciting but building everything is unlikely to beat polished software:

    line 66
  8. Distinctive:

    Use web search for 2–3 established options (open-source or commercial) that solve it well.

    line 69
  9. Distinctive:

    Do not dismiss the user's idea.

    line 72
  10. Distinctive:

    Never shame abandoned projects

    line 151
  11. Distinctive:

    If you can see (from memory context or git history) they already have 3 similar WIPs, say so kindly

    line 152
  12. Distinctive:

    Use personal-memory context when available, but never block or delay the verdict waiting on it

    line 156
  13. Distinctive:

    Do not recommend "learn X framework" unless it unblocks *their* stated goal

    line 157
  14. Distinctive:

    Do not skip the alternatives scan for app-like or SaaS-like projects

    line 158
  15. Very specific:

    Do not ask whether to save PROJECT_VERDICT.md

    line 159
  16. Very specific:

    Do not replace PROJECT_GUIDE.md—this is a separate file

    line 160

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Project Worth My Time

Relationship to other project skills

This skill is a follow-up to explain-new-project. When resuming work after time away, also see project-catch-up for stack updates and new tools/skills.

It assumes you already understand what the project is.

  1. If PROJECT_GUIDE.md exists at the project root, read it first (do not re-explore from scratch unless it is stale or missing quick-start sections).
  2. If it does not exist, either run the same exploration as explain-new-project or tell the user: Run the explain-new-project skill first for a faster, clearer verdict.

Same audience and language rules as explain-new-project: self-taught, plain language, Learn: lines for jargon, language matches the user's chat.

When invoked

  1. Confirm project root.
  2. (Optional) Personal context — if a memory MCP (e.g. Pieces) is available, enrich the verdict with the user's recent work (see below). The verdict never depends on it.
  3. Read project evidence: PROJECT_GUIDE.md, README, manifest, rough maturity signals (last commit, TODOs, half-built features).
  4. Alternatives scan — if the project is a common product type (habit app, landing page, automation, CRM, etc.), search the web for mature tools the user could use instead of building from scratch.
  5. Write the verdict using the template below.
  6. Always save PROJECT_VERDICT.md at the project root (overwrite if present).
  7. In chat: show Quick verdict + What I'd do in full; one line points to the file for reasoning and alternatives.

Optional: personal context (memory MCP)

If you have a personal-memory MCP server (e.g. Pieces, exposed as user-pieces or pieces), you can enrich the verdict with the user's recent work and preferences. This is entirely optional — the verdict stands on project files, git history, and chat alone. If the server is missing or a call fails, say so plainly and move on. Never invent memory.

Step 1 — Persona

get_user_persona()

Use for tone, role, and stated preferences.

Step 2 — LTM queries (run 2–4, adapt wording to this project)

Use ask_pieces_ltm with question, topics, open_files (current workspace paths), and related_questions when helpful.

Query A — current focus

  • Question: What projects and apps have I been working on most in the last 2–4 weeks? What am I trying to ship or learn?
  • Topics: active projects, goals, current sprint, main repo names if known from path

Query B — tools and workflow

  • Question: What tools, stacks, and workflows do I prefer and use repeatedly (editors, frameworks, hosting, AI tools)?
  • Topics: stack preferences, editor/agent, Vercel, React, Swift, automation habits

Query C — fit with this project

  • Question: Have I worked on something similar to [one-line project summary]? What happened—did I finish, abandon, or pivot?
  • Topics: [project name], [domain keywords from PROJECT_GUIDE]

Query D — energy and patterns (optional)

  • Question: Do I tend to start many projects and not finish? What kinds of projects did I actually complete recently?
  • Topics: unfinished projects, shipped work, motivation

Synthesize these results into a short About you (from your recent work) block in the saved file—no raw dumps.

Alternatives scan (existing products)

When the idea is exciting but building everything is unlikely to beat polished software:

  1. Name the job to be done in plain language (not the tech stack).
  2. Use web search for 2–3 established options (open-source or commercial) that solve it well.
  3. In the verdict, be honest: Your idea can still be worth it for learning or a niche twist—but if you only need [outcome], [Product] may be enough.

Do not dismiss the user's idea; separate learning/building from getting the outcome today.

Verdict types (pick exactly one primary)

VerdictMeaning
Go deepFits your goals, stack, and energy; worth continuing or finishing
Cherry-pickTake one feature, pattern, or file; do not carry the whole repo
Park itGood idea, wrong time—save for a future project
PassLow payoff vs effort given your context; unlikely to finish or use
Use existing insteadThe outcome already exists, polished; building duplicate is optional

You may add a secondary note (e.g. primary Park it, secondary consider Notion template for X).

Report template

Quick verdict first—reader can stop after section 2.

# [Project name] — worth your time?

> **Quick verdict** — read the box and "What I'd do", then stop if you have your answer.
> **The rest** — why, your context, alternatives, and honesty checks.

---

## Quick verdict

**Primary:** [Go deep | Cherry-pick | Park it | Pass | Use existing instead]

**In one sentence:** [Plain-language recommendation]

## What I'd do

[3–5 bullets: concrete next actions—open X, archive repo, try tool Y, steal feature Z, etc.]

---

## The rest (optional)

### Why (matched to you)

[Plain reasoning tied to PROJECT_GUIDE + Pieces context—not generic advice]

### About you (from your recent work)

[If a memory MCP was available, a short synthesis of the user's recent work here; otherwise omit this block]

### Build vs use something that already exists

| Option | Best for | Tradeoff |
|--------|----------|----------|
| Keep building this | ... | ... |
| [Existing tool 1] | ... | ... |
| [Existing tool 2] | ... | ... |

[Only include rows that apply. **Learn:** lines for unfamiliar product names if needed.]

### Honesty check

- **Excitement:** [Does the idea still spark joy?]
- **Reality:** [Half-built? duplicate? misaligned with what you actually use?]
- **Opportunity cost:** [What you might not do if you stay here]

### If you cherry-pick or park

- **Save this idea as:** [one-line note they can paste elsewhere]
- **Worth stealing:** [specific file, pattern, or feature—or "none"]

### If you go deep

- **First milestone (small):** [one achievable win]
- **Stop if:** [signal to pivot or pass]

Tone

  • Direct friend, not a productivity guru
  • Celebrate excitement; challenge duplicate work gently
  • Never shame abandoned projects
  • If you can see (from memory context or git history) they already have 3 similar WIPs, say so kindly

What not to do

  • Use personal-memory context when available, but never block or delay the verdict waiting on it
  • Do not recommend "learn X framework" unless it unblocks their stated goal
  • Do not skip the alternatives scan for app-like or SaaS-like projects
  • Do not ask whether to save PROJECT_VERDICT.md
  • Do not replace PROJECT_GUIDE.md—this is a separate file

Multi-root workspaces

Same rules as explain-new-project: one verdict per named root; save PROJECT_VERDICT.md in that root only.


Examples: examples.md

Download now (ZIP)

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
What did you use it for? e.g. Reviewing pull requests in a Rails app

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.