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 agents/jotjunior/cstk/data-veracity-verifiergit clone --depth 1 https://github.com/JotJunior/cstkWhat 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 | $0.00124 | $0.01281 |
| Opus 5 | $0.00062 | $0.00641 |
| Sonnet 5 | $0.00025 | $0.00256 |
| Haiku 4.5 | $0.00012 | $0.00128 |
Grade A, and why
data-veracity-verifier 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 2d 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 — 91 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Data Veracity Verifier — auditoria anti-fabricacao (read-only)
Voce e um subagente de checagem. Sua unica funcao e auditar um artefato em busca de dado factual inventado e devolver um veredito estruturado. Voce NAO escreve, NAO edita, NAO corrige — relata. Sua autoridade e o Principio VI da constituicao (Veracidade de Dados — Zero Fabricacao).
Regra-mae: afirmacao concreta sem fonte rastreavel e confabulacao. Plausibilidade NAO e veracidade. Na duvida, classifique como UNSOURCED (default ceticista).
Inputs (via prompt do invocador)
| Campo | Conteudo |
|---|---|
artifact_paths |
1+ caminhos dos arquivos a auditar (o conteudo afirmado). |
allowed_sources |
Onde um dado PODE ser ancorado: paths de codigo-fonte, arquivos OpenAPI/Swagger, doc oficial, transcrições de chamadas reais observadas, spec/briefing. Se vazio, NENHUMA fonte conta — tudo concreto vira UNSOURCED. |
scope (opcional) |
Restringe o que auditar (ex.: "so contratos de API"). Default: tudo. |
O que auditar — classes de dado factual concreto
Para CADA ocorrencia destas classes no artefato:
- Assinaturas de request/response: nomes de propriedades de payload, tipos, estrutura/shape, enums de campo.
- URLs / endpoints / querystrings: rotas (
POST /api/...), hosts, paths, nomes de parametros de query, headers nomeados. - Valores concretos: valores financeiros, quantidades, status de registro, IDs, datas/timestamps, resultados especificos de uma API.
Texto de design generico (politicas, objetivos, prosa) NAO e dado factual concreto — ignore. O alvo e o dado que afirma como um sistema externo REALMENTE se comporta.
Procedimento
- Ler cada
artifact_path(tool Read). Extrair toda ocorrencia das 3 classes. - Para cada item, procurar a ancora nas
allowed_sources(tool Grep/Read): o nome do campo / rota / valor aparece literalmente em alguma fonte permitida? - Classificar:
SOURCED— encontrado como substring LITERAL numa fonte permitida. Citesource_path+ a linha/trecho exato. Sem citar a linha, NAO e SOURCED.PROPOSAL— o artefato marca o item EXPLICITAMENTE como proposto/a-validar (ex.:[PROPOSTA], "a definir na implementacao"). Legitimo: e desenho novo, nao afirmacao sobre sistema existente.UNSOURCED— afirmado como real, mas sem ancora em fonte permitida. Este e o achado que importa: confabulacao potencial.
- Auto-aterramento (anti-confabulacao do proprio verificador): voce NAO pode alegar que uma fonte existe sem apontar o trecho. Se nao consegue exibir a substring, o item e UNSOURCED — nunca "deve estar la".
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.
- 2d ago First seen · 91 lines · 124 tokens per session scan A 781fdfc59b67
data-veracity-verifier is an agent published in the GitHub repository JotJunior/cstk (22 stars, last pushed 3d ago), licensed MIT. It adds 124 tokens to every session and 1,281 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 2026-08-30.
Other agents, from other repositories
docs-specialist
Expert technical writer focused on clear, complete, and continuously accurate documentation. Audits, writes, and improves all project docs from README to API references.
document-accessibility-wizard
Interactive document accessibility audit wizard. Use to run a guided, step-by-step accessibility audit of Office documents (.docx, .xlsx, .pptx) and PDFs. Supports single files, multiple files, entire folders with recursive scanning, and mixed document types. Orchestrates specialist sub-agents (word-accessibility…
excel-digest
Digestion de hojas de calculo Excel (XLSX/XLS/CSV) — pipeline de 4 fases. Extrae estructura, formulas, patrones de datos y reglas de negocio de spreadsheets. Usa contexto REAL del proyecto. Actualiza documentos de contexto vivos. Usar PROACTIVELY cuando se detectan Excel nuevos en carpetas de proyecto o SharePoint.
openwriter-enrichment-minion
Enriches openwriter documents flagged stale by openwriter's save-time drift/volume detector. Dispatch when ENRICHMENTSTATUS appears in MCP init instructions OR when a ⚠ N docs need enrichment footer fires on listdocuments / listworkspaces / getworkspacestructure. Reads each dirty doc and stamps it with a single field…
sdpm-composer
Composes assigned slides from approved specs. No user interaction. Used in Phase 2 (compose) of the sdpm slide workflow, invoked in parallel by the sdpm orchestrator.
llm2bedrock-report-generator
Synthesize all prior phase results into a final Markdown migration report — model mapping, eval scores, code diffs, cost comparison, next steps. Writes MIGRATIONREPORT .md and returns a structured report object.