agentleFS
Sign inStart free
KemetraKemetra/Seshat-BI/CLAUDE.md

Seshat-BI

CLAUDE.md2 starsChanged 4 months agoApache-2.0View on GitHub

Repo-specific rules. Global rules in ~/.claude/CLAUDE.md still apply. A standalone analytics service — NOT bound by the Retail Tower OS orchestrator / contract-boundary rules. Power BI primary; DigitalOcean Postgres source. Data flows bronze → silver → gold; Power BI reads the gold schema only. For new retail mart work, start from a filled worked example under docs/worked-examples/ and follow the medallion playbook. - Secrets: credentials only in .env (git-ignored). Never write real values into tracked files. Power BI uses parameters,…

  • Reads credentials

The file

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

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

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈2

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    No live DB provisioning, no automated ingestion code, no orchestrator integration unless explicitly requested.

    line 33
  2. Distinctive:

    No active Spec Kit implementation plan.

    line 37

≈ “Common” is an estimate: at least 1 in 5 of the 26,980 CLAUDE.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
CLAUDE.md

CLAUDE.md — Seshat BI

Repo-specific rules. Global rules in ~/.claude/CLAUDE.md still apply.

What this repo is

A standalone analytics service — NOT bound by the Retail Tower OS orchestrator / contract-boundary rules. Power BI primary; DigitalOcean Postgres source. Data flows bronze → silver → gold; Power BI reads the gold schema only.

For new retail mart work, start from a filled worked example under docs/worked-examples/ and follow the medallion playbook.

Hard rules

  • Secrets: credentials only in .env (git-ignored). Never write real values into tracked files. Power BI uses parameters, not baked-in connection strings.
  • PBIP .gitignore baseline is exact: **/.pbi/localSettings.json and **/.pbi/cache.abf. Never ignore definition/ folders — that's the model.
  • PBIP is a preview feature (as of 2025-12); enable it in Power BI Desktop.
  • Windows 260-char path limit — keep PBIP project/table names short.
  • Line endings: core.autocrlf=true; rely on .gitattributes. Edit PBIP text externally only as UTF-8 without BOM.

Conventions

SQL: snake_case; schemas bronze/silver/gold; vw_/fct_/dim_ prefixes; numbered idempotent migrations. DAX: PascalCase measures in display folders. Full detail in docs/conventions.md.

Scope discipline (YAGNI)

No live DB provisioning, no automated ingestion code, no orchestrator integration unless explicitly requested. Add the seam, not the implementation.

<!-- SPECKIT START -->

No active Spec Kit implementation plan.

<!-- SPECKIT END --><!-- SESHAT-KIT START -->

Seshat BI kit router (v0.2.0) -- generated from .seshat/kit-source.yaml; do not edit here.

Orient first: What readiness stage am I serving? State lives in readiness-status.yaml (per TABLE, recomputed) (recomputed; this file stores none).

Verbs the agent drives:

  • retail-orchestrate -- conductor -- sequence the medallion verbs, self-heal against the gate
  • first-hour-compass -- first-arrival worked-example offer + single-source seam list + single-table orientation card
  • retail-onboard-table -- Source->Mapping front door; owns the Stage-1 read-only DB-backed profile (grain candidates, column types)
  • retail-discover-portfolio -- metadata-only portfolio discovery -> governed domain/scope proposals -> selected-table onboarding -> interview handoff
  • business-knowledge-interview -- after DB discovery, interview the owner into the Decision Store (batch low-risk, explicit critical); records decisions, never self-grants approval
  • source-mapping -- the mapping gate -- produces source-map.yaml
  • kpi-contract-builder -- drive the shipped kpi_contracts engine: assess answerability, list the decisions to approve, preview with per-field provenance, then draft/finalize -- never self-grants approval
  • retail-build-warehouse -- author silver/gold SQL; stop before executing
  • retail-validate -- live checks; needs db extra + DSN, else [PENDING LIVE PROFILE]
  • retail-govern -- static check (seshat check)

Hard-stops (orientation the agent reads; enforcement is the lint rules + G6/C2, not this file):

  • never_self_grant_approval
  • no_silver_before_mapping_cleared
  • no_dashboard_before_metric_contracts
  • never_fabricate_a_confidence_score
<!-- SESHAT-KIT END -->

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