agentleFS
frankxaifrankxai/second-brain-os/AGENTS.md

second-brain-os

AGENTS.md0 starsChanged 42 days agoMITView on GitHub

Persistent personal intelligence and cognitive memory operating system. Built on the Starlight Intelligence Protocol (SIP) Layer 1. This file has two audiences: (1) coding agents contributing to this repo, and (2) the product's own agents that run inside an SBO vault. Read the section relevant to your task. Two OSS starter agents (below) plus a paid tier of 8. This repo is the OSS template only — the paid agents live in the separate @frankx/second-brain-pro package. See docs/paid-tier.md for the…

  • Installs packages
  • Where files are headingNew AGENTS.md files have more rules about keeping secrets out of the code, and fewer about code style.

    Of the AGENTS.md files we read, the 33,315 first seen in the last six months, against the older ones: the kind of rule that gained most share, and the kind that lost most. About files like this in general, not this one, and an estimate, since a kind is a pattern in the words.

  • In this repository4 other files it gives its agents.

The file

We split this AGENTS.md into the rules it gives an agent

8 rules. About 0 are common, 3 are distinctive, and 5 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈3

    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:

    Voice gate: no AI-slop phrases in docs or generated content (d-e-l-v-e, d-i-v-e into, etc.) — enforced by src/sbo_ingestion/voice_check.py …

    line 19
  2. Distinctive:

    All raw notes and incoming signals follow:

    line 58
  3. Very specific:

    Automate: Run the OSS starter agents (people-map, pattern-detector) or optional paid-tier agents (@frankx/second-brain-pro).

    line 62
  4. Very specific:

    Weekly (OSS): Run people-map (/people-update) and pattern-detector (/patterns-detect).

    line 66
  5. Very specific:

    Monthly (Paid tier — @frankx/second-brain-pro): Run business-map, decision-history-miner (publishes to _meta/decisions-history.md), and …

    line 67
  6. Very specific:

    Quarterly (Paid tier — @frankx/second-brain-pro): Run psychometric (16p-mapper, stp-mapper, enneagram-analyzer) and values …

    line 68
  7. Distinctive:

    All agents emit SIP attestation blocks on artifact creation:

    line 79
  8. Distinctive:

    When motion, scroll, generated media, GIF/video, or premium polish matters, route through the Motion Design Studio plugin/skills and verify …

    line 99

≈ “Common” is an estimate: at least 1 in 5 of the 42,283 AGENTS.md 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
AGENTS.md

Second Brain OS — Agents & Intelligence Architecture

Persistent personal intelligence and cognitive memory operating system. Built on the Starlight Intelligence Protocol (SIP) Layer 1. This file has two audiences: (1) coding agents contributing to this repo, and (2) the product's own agents that run inside an SBO vault. Read the section relevant to your task.

Working in this repo (for coding agents)

  • Branch: main is the trunk. Cut a feature branch per change (agent/<harness>/<scope> if you're one of several parallel agents — see the estate-wide multi-agent protocol in the parent starlight/repos/CLAUDE.md). Never commit directly to main from an automated session unless told to.
  • Install: python -m venv .venv && source .venv/bin/activate (or .venv\Scripts\activate on Windows), then pip install -e ..
  • Test: pytest -v must be green before you open a PR. The suite covers the Claude.ai/ChatGPT handlers, the summarizer (mocked Anthropic), voice check, dual-write, the distill CLI, end-to-end ingest across all three modes, and an adversarial suite (path traversal, NTFS streams, prompt injection, malformed timestamps). CI (.github/workflows/test.yml) runs the full matrix on ubuntu/macos/windows x Python 3.11/3.12/3.13.
  • Privacy gate: if you touch scripts/, templates/, or src/sbo_ingestion/dual_write.py, also run bash scripts/verify-privacy.sh templates/private-vault-skeleton (or the .ps1 on Windows) and confirm it exits 0 with zero failures. Machine-level items (Windows Search exclusion, macOS Time Machine exclusion) report WARN with their manual remediation and do not fail the run. This is the check that the private/ vault stays air-gapped from MCP.
  • Voice gate: no AI-slop phrases in docs or generated content (d-e-l-v-e, d-i-v-e into, etc.) — enforced by src/sbo_ingestion/voice_check.py and checked in CI.
  • Non-negotiable invariant: nothing you add may give MCP (or any LLM connector) a path into private/. The two-vault boundary is filesystem-level, not config-level — don't "fix" it with a permission flag.
  • Full PR checklist: see CONTRIBUTING.md.
  • Don't touch: maintainer-local harness state (.asph-wip/, .grok/, .agent-harness.json) is gitignored. If it reappears in git status, ignore it rather than staging it.

What ships in this template

Two OSS starter agents (below) plus a paid tier of 8. This repo is the OSS template only — the paid agents live in the separate @frankx/second-brain-pro package. See docs/paid-tier.md for the OSS/paid line.

OSS starter agents (free)

people-map

  • Trigger: /people-update slash command (or scheduled weekly via Starlight Chronicle).
  • Reads: brain/_inbox/**/*.md, brain/notes/**/*.md, brain/projects/**/*.md.
  • Writes: brain/people/{slug}.md (one file per person mentioned).
  • Contract: see _agents/people-map.md in any brain vault.

pattern-detector

  • Trigger: /patterns-detect slash command (weekly via Chronicle).
  • Reads: brain/notes/, brain/_inbox/ summaries (last 30 days).
  • Writes: brain/patterns/YYYY-Www.md.
  • Contract: see _agents/pattern-detector.md.

Paid tier agents (FrankX premium)

Installed via @frankx/second-brain-pro npm package on purchase. Not in this OSS template.

  1. big5-inferer — Big 5 personality inference. Monthly. Writes _meta/psychometrics/big-5.md. Ships with mandatory validity disclosure.
  2. 16p-mapper — MBTI/16P typology. Quarterly. Validity disclosure.
  3. stp-mapper — StrengthsFinder analog. Quarterly. Validity disclosure.
  4. enneagram-analyzer — Enneagram typology. Quarterly. Validity disclosure.
  5. business-map — extracts businesses, ventures, projects. Monthly. Writes _meta/businesses.md.
  6. decision-history-miner — pulls decisions from notes/decisions/ + _inbox/. Monthly. Writes _meta/decisions-history.md.
  7. ikigai-extractor — values + ikigai analysis. Quarterly. Writes _meta/values.md.
  8. content-engine-integration — feeds the FrankX content pipeline. Weekly.

Operational Execution Layer & Rituals

The 4-Step Ingestion Cycle

All raw notes and incoming signals follow:

  1. Drop: Unfiltered raw capture into brain/_inbox/ with zero cognitive friction.
  2. Extract: Semantic atomic distillation into brain/notes/ with explicit tags and references (sbo-distill).
  3. Direct: Connect to brain/projects/, brain/people/, or decision notes.
  4. Automate: Run the OSS starter agents (people-map, pattern-detector) or optional paid-tier agents (@frankx/second-brain-pro).

Ritual Cadence

  • Daily (OSS): Git health sync; run sbo-distill to triage brain/_inbox/ into atomic notes.
  • Weekly (OSS): Run people-map (/people-update) and pattern-detector (/patterns-detect).
  • Monthly (Paid tier — @frankx/second-brain-pro): Run business-map, decision-history-miner (publishes to _meta/decisions-history.md), and big5-inferer.
  • Quarterly (Paid tier — @frankx/second-brain-pro): Run psychometric (16p-mapper, stp-mapper, enneagram-analyzer) and values (ikigai-extractor) baseline updates.

Planned Multi-Agent Ecosystem Integrations (Roadmap)

The following ecosystem integrations live in separate Starlight repositories and are planned opt-in bridges rather than bundled dependencies of this OSS template:

  • SIS (Layer 0 Substrate): Optional cross-vault sync with the 6 core Starlight vaults (Strategic ◆, Technical ⬡, Creative ✦, Operational ▸, Wisdom ◎, Horizon ↗).
  • Hermes Agent: Optional cron-scheduled maintenance and multi-source synthesis with provenance.
  • OpenClaw: Optional sovereign security boundary and cryptographic attestation.

Built on SIP

All agents emit SIP attestation blocks on artifact creation:

**SIP Attestation**
protocol: SIP v1.1.1
analyzed_at: YYYY-MM-DD
sources: Second Brain vault artifacts

Validity disclosures are non-removable.

Design Taste Kernel

For any site, app, landing page, dashboard, visual identity, brand, motion, media, social, or frontend task, apply the shared Design Taste Kernel (design-agent-standards) before handoff:

  • DESIGN_TASTE.md
  • WEB_EXPERIENCE_STANDARD.md
  • MOTION_TASTE_RUBRIC.md
  • MULTI_AGENT_DESIGN_COUNCIL.md
  • VISUAL_QA_GATE.md

When motion, scroll, generated media, GIF/video, or premium polish matters, route through the Motion Design Studio plugin/skills and verify the result visually.

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.