agentsclimarketplace

Customs validate

Skill FrancoisChastel/sydonia-toolkit/skills/customs-validate

Write friendly SQL, compile it to genuine ASYCUDA World (SYDONIA) customs SQL you can run read-only. A fully-sourced PostgreSQL reference model + query compiler + docs + Agent Skills for customs analytics, ML & selectivity.

Install
npx -y skills add FrancoisChastel/sydonia-toolkit --skill customs-validate

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

2 things to look at

  • 21 days oldThe repository was created 21 days ago. New is not bad, but a brand new repository carrying a familiar-sounding name is the shape a typosquat arrives in, and there has been no time for anyone else to find a problem with it.
  • 1 stars1 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Validate, verify or audit the Sydonia Toolkit customs schema against its done-conditions — a clean load (schema + seed + end-to-end example, zero errors), every table carrying a provenance tag, and every cited source ID resolving in SOURCES.md. Use when the user wants to check the schema still loads cleanly, confirm it is fully grounded/sourced, or audit a fork after changes before committing.

SKILL.md

2.7 KB, 611 tokens by cl100k_base, as published. Nobody here has run it

Customs validate

Re-run the project's four done-conditions and report a PASS/FAIL. This is the auditable green light after any change, and the honest check before a commit.

When to use

Trigger on "validate / verify / audit / check the customs schema", after using customs-extend or customs-seed, or whenever someone asks "is it still clean and fully sourced?".

How to run

The bundled script (relative to this skill's folder) does everything against a throwaway database and cleans up after itself:

bash scripts/verify.sh [--sydonia-dir DIR] [--db NAME]
  • Auto-detects the Sydonia/ directory from the project root; pass --sydonia-dir if it lives elsewhere.
  • Requires psql + createdb (PostgreSQL 14+). If absent, it still runs the static checks (provenance tags, source resolution) and clearly reports that the load check was skipped.

The four checks

  1. Schema loads clean. asycuda.sql → seed_reference.sql → examples/e2e.sql against a fresh database with ON_ERROR_STOP=1, zero errors, and the e2e example inserts (assessed total equals the receipt).
  2. Every table is grounded. Every CREATE TABLE in schema/*.sql has a preceding -- src: <ID> or -- inferred. Zero untagged tables.
  3. Every source resolves. Each <ID> cited in the schema has a row in SOURCES.md. (The script also flags IDs with no cached file under sources/ or docs/.)
  4. Docs present. DATA_DICTIONARY.md, ERD.md, COVERAGE.md, RESEARCH_LOG.md exist.

Interpreting the result

  • All PASS → report the table count (expect 55), the documented/inferred split, and the assessed=receipt line. Safe to commit.
  • Any FAIL → surface the exact failure (the untagged table, the unresolved ID, or the psql error) and stop. Do not "fix" a failure by loosening a check or inventing a citation — that defeats the purpose. If a new table is genuinely unsourced, it must be tagged -- inferred and noted in COVERAGE.md (use the customs-extend skill).

Notes

This mirrors the acceptance checks in Sydonia/SYDONIA-GOAL.md §1. Keep them honest: a larger truthful inferred set is always preferred over a fabricated documented one.

Keep looking

Skills are one crate of 328,083. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.