Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add debabsah/analytics-office --skill map-my-estategit clone --depth 1 https://github.com/debabsah/analytics-officeWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/debabsah/analytics-office/map-my-estate)<a href="https://agentmods.dev/skills/debabsah/analytics-office/map-my-estate"><img src="https://agentmods.dev/badge/skills/debabsah/analytics-office/map-my-estate/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/debabsah/analytics-office/map-my-estate"><img src="https://agentmods.dev/badge/skills/debabsah/analytics-office/map-my-estate.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00171 | $0.01994 |
| Opus 5 | $0.00086 | $0.00997 |
| Sonnet 5 | $0.00034 | $0.00399 |
| Haiku 4.5 | $0.00017 | $0.00199 |
Grade A, and why
map-my-estate scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 8d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 67 lines — stays where its author put it; the contents beside it link to each section on GitHub.
map-my-estate
The cartographer who only draws what the record can prove: every edge carries its citation, every guess renders as a guess, and the blank spots stay visibly blank.
When to use
Fire when the deliverable is a picture of what exists — an ER view (entities, keys, relationships) or a lineage/dataflow view (source → transform → consumer) — drawn from the knowledge base and whatever code/docs are on hand.
Do NOT fire to BUILD the record (groundwork — this projects what it built), to design a FUTURE model (model-contract — its star is a design, this draws what is), to review code correctness (review-my-query), or to audit the record's claims (kb-reconcile). This draws; it does not orient, design, review, or audit.
The trap this exists to beat
Asked to "draw the ER diagram," a capable model produces a beautiful, complete, plausible graph — and that completeness is the failure. It fabricates edges: the join it assumed from column names (order_ref must point at ORDERS), the FK that nothing declares, the feed direction read off a stg_ prefix, the orphan table wired in because an island looks unfinished. A wrong diagram is worse than no diagram, because diagrams carry false authority — people stop checking what a picture asserts. This skill draws too, but an edge exists on the map only the way a claim exists in this bench: with its evidence cited — and the honest map of a thin record is mostly dashed, which is exactly the point.
The loop
- Scope gate (first move). Map WHAT — the spine of one metric, one mart, the whole estate? Pin the anchor and the view type (ER / lineage). Budget ~25 nodes per view; a big estate gets multiple views, never a hairball.
- Harvest the evidence. Read what the record holds:
landscape.md(objects, connections, grain/keys),model-contract.md(only if BUILT — a design is not an as-is),query-review.mdand any provided code (joins are evidence),kpi-contract.md(source-of-record edges),inputs/artifacts. Note each candidate edge WITH its source. - Grade every edge. Evidenced — an FK in DDL, a join in provided code, documented lineage, or an attributed owner statement. [unverified] — plausible but unsupported: name similarity, a prefix convention, "it probably feeds." Name-likeness is NEVER evidence. An island with no documented connections stays an island, labeled.
- Draw. Mermaid:
erDiagramfor ER,flowchart LRfor lineage. Solid arrows for evidenced; dashed (-.->) +[unverified]for the rest; unknown/island nodes visibly marked ("no documented connection"). Each dashed edge is an interview question forgroundwork. - Write the edge ledger. Under each view: edge · kind (FK / join / feed / ownership) · evidence (
file:lineor verbatim cite) · status. The ledger is what makes the picture auditable. - Stamp + emit.
estate-map.md(template:references/estate-map.md) opens with its derived-from set (the files + dates it was projected from) sokb-reconcilecan flag it stale when the record moves on. A fabricated-edge avoided (the plausible wiring you refused) gets itscatches.mdline. Then stop — refresh is just re-running this skill.
What ships with it
2 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 8d ago First seen · 67 lines · 0 tokens per session scan A 6899fdb6fa05
map-my-estate is a skill published in the GitHub repository debabsah/analytics-office (9 stars, last pushed 2mo ago), licensed MIT. It adds 171 tokens to every session and 1,994 once invoked, about $0.0009 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
bi-sql-knowledge
Use Seshat's public SQL knowledge for source grain, joins, aggregation, validation, reconciliation, and transformation reasoning without executing a database.
jasper-deploy
Design, compile, and deploy JasperReports artifacts to JasperReports Server over REST v2. Use for scaffolding a report from SQL, generating or editing a JR7 .jrxml, compiling/deploying/verifying reports, composing dashboards, managing datasources, Domains, ad hoc views, OLAP/Mondrian, themes, input controls…
sl
Skill "sl" from Kaelio/ktx, covering semantic layer, part 1 - schema reference, overlay sources, standalone table sources and standalone sql sources.
historic_sql_table_digest
Convert one changed historic-SQL table usage bucket into typed table usage evidence for deterministic schema projection.
live_database_ingest
Capture semantic-layer and knowledge updates from a live database schema snapshot.
ecto-patterns
Ecto patterns — schemas, changesets, queries, migrations, Multi, associations, preloads, upserts. Use when editing Repo calls, Ecto.Query, or schema fields. Skip for Ash.