# Fixture Ledger Lookup

> Fictional command-line lookup over an embedded company ledger snapshot.

- Skill: `firatcand/fixture-ledger-lookup` (Agent Skill, multi-file: 2 files)
- Install (CLI): `npx skillmds@latest add firatcand/fixture-ledger-lookup`
- Raw SKILL.md: https://api.skillmd.com/api/skills/firatcand/fixture-ledger-lookup/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: firatcand (https://skillmd.com/u/firatcand)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/firatcand/fixture-ledger-lookup

---


# fixture-ledger: lookup

A fictional, deterministic command-line surface. This skill owns the entire
provider contract — invocation syntax, input schema, output schema, and
authentication mechanics. Workspace tool-use definitions carry only company
policy and never restate anything on this page.

## Invocation

Run the adjacent `invoke.mjs` executable with exactly one JSON argument passed
as a literal argv value, never through shell string interpolation:

```
node invoke.mjs '{"account_code":"...","ledger":"...","exclude_flags":[],"fields":["..."]}'
```

The executable reads no environment variables, opens no sockets, spawns no
processes, and reads no files. Its whole dataset is embedded.

## Input

- `account_code`: one lowercase account identifier.
- `ledger`: the single ledger partition the caller may read.
- `exclude_flags`: account flags the caller must never see.
- `fields`: the exact output fields the caller requires per result.

## Output

One JSON document on stdout: `{"results":[...]}`. Each result carries exactly
the requested `fields`. Exit code 0 on success, 2 on malformed input.

## Authentication

None. This fixture is credential-free by construction. A real deployment would
resolve ambient credentials outside the workspace; credential material never
belongs in authored workspace files.