Your first Praxis Mine result

This walkthrough takes one candidate from vague interest to a bounded next move. It does not install or execute the candidate.

Choose a candidate

Bring one outside AI skill, repository, tool, prompt pack, or workflow. You need:

• its name and link or local path;
• the outcome you hope it improves;
• what you already use for that outcome;
• any hard limits on privacy, security, license, money, runtime, or dependencies.

Sparse input is acceptable. Praxis Mine should identify the missing evidence that changes the disposition instead of making you complete an intake form from the Ministry of Forms.

Ask for the mine

Use:

Assess this outside skill and steal only the good bits. I want it to improve [outcome]. My current baseline is [baseline]. The candidate is [link, path, or description]. My hard limits are [limits].

If the host cannot access the link or path, attach or paste the smallest relevant files. Do not provide credentials or private material merely to make the analysis look complete.

Inspect the result

Require six elements:

1. Disposition: pilot, quarry, adapt, monitor, or reject.
2. Capability delta: the outcome that becomes possible, better, cheaper, faster, or safer.
3. Evidence: what was observed, supplied, inferred, or remains unknown.
4. Costs and risks: dependencies, permissions, privacy, rights, maintenance, money, context, and overlap.
5. Baseline comparison: why this beats—or fails to beat—what you already have.
6. Next action: a small falsifiable check with a clear result when more evidence is needed, or the reason to stop for a justified skip.

If the answer says adopt without a separate acceptance gate, or recommends installation because a package is popular, the result failed Praxis Mine’s governing boundary.

Optional: create a local ledger

Open a terminal in the installed skills/praxis-mine directory:

python scripts\praxis_mine.py --data-home .\praxis-data init
python scripts\praxis_mine.py --data-home .\praxis-data ingest-manifest assets\seed-manifest.json
python scripts\praxis_mine.py --data-home .\praxis-data status

Expected results:

• init reports a healthy praxis_mine store;
• ingest-manifest creates ten seed candidates on the first run;
• repeating the ingest reports those candidates unchanged;
• status reports candidate, evaluation, and awaiting-evaluation counts.

The .\praxis-data directory is disposable first-run data. See Ledger reference before choosing a durable location.

You are done when

For an assessment request, you can explain what is worth keeping and why the evidence warrants the decision. For an authorized adaptation request, continue through the actual capability owner to the verified change. The local room records assessments and never installs or executes a candidate by itself.

The bundled seed scores are historical example inputs, not current verified product recommendations. They remain usable for learning the ledger, but lack the new declared comparison/rights basis and therefore cannot earn an actionable disposition without actual reassessment.