agentleFS
Sign inStart free
tonydzitonydzi/second-brain-starter-kit/skills/content-mine/SKILL.md

content-mine

Skill6 starsChanged 2 months agoMITView on GitHub

Scan recent agent sessions for content-worthy moments and capture them as drafts in a publishing funnel, publishing nothing. Zero-token detector, then a taste judge, then capture. Triggers: "/content-mine", "mine sessions for content", "what in our sessions deserves a post".

1 other project uses a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

2 rules. About 0 are common, 1 is distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 1

    Very specific

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

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

    READ the digest cand-latest.md (or OUT).

    line 25
  2. Distinctive:

    Don't build a parallel monolith — reuse content_miner + voice_triage + /episode.

    line 34

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 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

CONTENT_MINE — sessions → funnel drafts (draft-first, taste delegated to me)

GOAL: mine our Claude Code sessions for content-worthy MOMENTS and drop them as DRAFTS into the content-factory funnel. Publishing happens only on the operator's explicit "+" through /episode. Nothing goes out.

The operator delegated the taste call — "what deserves to be content" — to me (2026-07-08). I don't ask what to add; I decide.

PATHS:

  • Engine: $IMPORTS_ROOT/content-factory\content_miner.py (same family as alpha-extraction / intention_mine; reads the live vault_sessions pool).
  • Funnel (intake): content-factory\triage\posts.jsonl (+ posts.md), source_kind: session. Writes through voice_triage.py append (dedup/upsert).
  • Digest: content-factory\miner\candidates\cand-latest.md. Ledger: miner\captured.jsonl.

STEPS:

  1. SCAN (deterministic, 0 tokens). Recent by default: python content_miner.py mine --days 7. Backlog / the whole archive: mine --all --operator Anton --cap 150. Flags: --day YYYY-MM-DD, --operator all, --cap N. Note OVER_THRESHOLD and OUT.
  2. Empty (SESSIONS=0 / OVER_THRESHOLD=0) → say "no fresh moments worth capturing" and stop.
  3. READ the digest cand-latest.md (or OUT). Each ### = a session src | date | machine | operator [PRIV] {score · tags} + a headline + snippets ([U]=the operator, [A]=Claude). Look at the top ones (HOT→WARM); on large volumes, fan out Sonnet judges over slices (see miner\slices\), they return JSON verdicts, you dedup and capture.
  4. JUDGE BY TASTE: keep the genuinely worthy and DIFFERENT stories (built/shipped it · a war with a bug / a root cause · a human+AI, multi-machine or consensus technique · a sharp insight · a "wow"). Drop routine, thin, forked duplicates (similar headline → one story), purely private material. Positioning: non-coder + AI is the trump card; the goal is an offer from an LLM company plus an audience of builders.
  5. CAPTURE the worthy ones: python content_miner.py capture --src <cc:xxxxxxxx> --title "<hook>" --note "<the gist>" --tier teaser|medium|longread|dev-log --angle "<angle>" --visibility public|personal|private. Tier: teaser=a hook; medium=the public diary post; longread=a narrative; dev-log=dry and technical. [PRIV] / anything that can't be scrubbed → --visibility personal. The engine blocks secret-looking tokens and dedups by src.
  6. BRIDGE INTO THE CANON (the law "an event → first a beat → then a post", [[show-canon-single-source]]): for the 1-2 strongest public moments create a draft BEAT in 04-Projects\show-canon\beats-inbox\ using the template beats\_TEMPLATE-beat.md (+authored_by: content-miner); never write into beats\ directly — the canon writer moves them. Older (pre-canon, before 2026-07-08) events get a backfill beat only when an episode is actually being assembled from them.
  7. REPORT to the operator (ELI5): how many sessions were scanned → how many drafts went into the funnel (with tiers) → how many beat candidates landed in the canon inbox; what's next = /episode on a "+", the outbound queue = registry\pub_registry.py (queue → next → posted, limits in limits.json). Show python content_miner.py captured --limit N.

BOUNDARIES (keep it simple and repairable):

  • Draft-first, HARD: drafts only; no auto-publish; nothing goes out.
  • Secrets/private material/CRM/leads never become content (engine gate + [[credential-store]]).
  • Don't build a parallel monolith — reuse content_miner + voice_triage + /episode.
  • Grunt work (detection/classification) = Sonnet; the authored post text = Opus (and that happens in /episode, not here).
  • This is the manual twin of the nightly content-miner-nightly; the archive backlog was closed in one pass on 2026-07-09 (miner\debt-log.md).
<!--kit-footer-->

Like this skill? It is one of 100 in second-brain-starter-kit: the second brain we built for ourselves and run every day at Palo Alto AI Research Lab. Install the whole set with npx skills add tonydzi/second-brain-starter-kit. Everything is open source and free, so take what you need.

Flagships worth a look on their own: secondop-panel (a second opinion from a panel of external models), claude-memory-tidy (stop your agent's memory from rotting), telegram-mcp-kit (your own Telegram over MCP in about 15 minutes).

Author: Anton Dziatkovskii, Palo Alto AI Research Lab. Telegram @tonydzi - WhatsApp +1 341 222 9178 - X @Tony_Stef_

Engineers: want to test-drive this setup? Message me. I hand out free starter seeds to engineers who test and report back, and custom skill requests are welcome.

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.

Check your setup, free