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 agentmods add skills/knuckles-team/epistemic-graph/kg-modality-sqlnpx skills add Knuckles-Team/epistemic-graph --skill kg-modality-sqlgit clone --depth 1 https://github.com/Knuckles-Team/epistemic-graphWrote 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/knuckles-team/epistemic-graph/kg-modality-sql)<a href="https://agentmods.dev/skills/knuckles-team/epistemic-graph/kg-modality-sql"><img src="https://agentmods.dev/badge/skills/knuckles-team/epistemic-graph/kg-modality-sql.svg" alt="Measured on agentmods" 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 | $0.00111 | $0.00801 |
| Opus 5 | $0.00056 | $0.00400 |
| Sonnet 5 | $0.00022 | $0.00160 |
| Haiku 4.5 | $0.00011 | $0.00080 |
Grade A, and why
kg-modality-sql 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 5d 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 — 64 lines — stays where its author put it; the contents beside it link to each section on GitHub.
kg-modality-sql — SQL over the engine wire
The engine speaks SQL over the actual wire protocols (DataFusion-backed eg-query,
wire-neutral classify → dispatch → exec core). Full SELECT (joins, aggregates,
GROUP BY/HAVING, window frames, CTE, subquery, UNION), INSERT/UPDATE/DELETE (with
compound WHERE, RETURNING, INSERT … SELECT, ON CONFLICT upsert), CREATE/ALTER/DROP
DDL on arbitrary durable user tables, views, and SQL/plpgsql functions — plus Postgres
extension drop-ins (pgvector <->/<=>, Apache AGE cypher(), TimescaleDB, ParadeDB BM25).
pg_catalog/information_schema are synthesized so psql \d, ORMs and BI tools introspect.
See docs/capabilities.md → SQL and Postgres wire.
The wire way (epistemic-graph owns it)
Each wire is opt-in (build with its feature + set its _ADDR env var). Postgres wire
(feature pgwire, EPISTEMIC_GRAPH_PGWIRE_ADDR, default 127.0.0.1:5433):
psql "host=127.0.0.1 port=5433 user=agent dbname=__commons__"
SET graph = 'my_graph';
SELECT id, properties FROM nodes LIMIT 10;
INSERT INTO nodes (id, properties) VALUES ('n1', '{"label":"Doc"}');
Simple and extended/prepared protocol; SCRAM-SHA-256 auth (the pg user becomes the ACL
actor, so RLS applies) or trust in dev. Sibling wires share the same SQL core:
MySQL (mysql-wire, :3306), MSSQL/TDS (mssql-wire, :1433), SQLite NDJSON
(sqlite-wire). Recipes + ports: docs/interfaces/connecting.md.
The MCP way (through graph-os)
load_tools(tools=["engine_query"]) # then run SQL through engine_query
or the REST twin graph-os exposes for the query modality.
Cross-modal seam
The nodes table is the same store the RDF, vector, and time-series modalities read, so a
SQL statement can JOIN graph ⋈ vector ⋈ timeseries and commits inside the engine's unified
mixed-store wire transaction (BEGIN/COMMIT/ROLLBACK, read-your-own-writes) —
cross-modal ACID, not a separate SQL database.
Related
kg-modality-sparql— RDF/SPARQL over the same nodes;kg-modality-consensus— the multi-Raft/tenant substrate these wires commit into.
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.
- 5d ago First seen · 64 lines · 111 tokens per session scan A 8a33d1994636
kg-modality-sql is a skill published in the GitHub repository Knuckles-Team/epistemic-graph (10 stars, last pushed 7d ago), licensed MIT. It adds 111 tokens to every session and 801 once invoked, about $0.0006 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
cdc
Change Data Capture - architecture, entrypoints, bytecode emission, sync engine integration, tests.
async-io-model
Explanations of common asynchronous patterns used in tursodb. Involves IOResult, state machines, re-entrancy pitfalls, CompletionGroup. Always use these patterns in core when doing anything IO.
differential-fuzzer
Information about the differential fuzzer tool, how to run it and use it catch bugs in Turso. Always load this skill when running this tool.
storage-format
SQLite file format, B-trees, pages, cells, overflow, freelist that is used in tursodb.
mvcc
Overview of Experimental MVCC feature - snapshot isolation, versioning, limitations.
transaction-correctness
How WAL mechanics, checkpointing, concurrency rules, recovery work in tursodb.