Borrowing it
Nothing to install: this file belongs to paultyng/testagent. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/paultyng/testagent/main/.claude/skills/verify-upstream-config-examples/SKILL.mdgit clone --depth 1 https://github.com/paultyng/testagentWrote 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/paultyng/testagent/verify-upstream-config-examples)<a href="https://agentmods.dev/skills/paultyng/testagent/verify-upstream-config-examples"><img src="https://agentmods.dev/badge/skills/paultyng/testagent/verify-upstream-config-examples/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/paultyng/testagent/verify-upstream-config-examples"><img src="https://agentmods.dev/badge/skills/paultyng/testagent/verify-upstream-config-examples.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 | $0.00029 | $0.01263 |
| Opus 5.5 | $0.00012 | $0.00505 |
| Sonnet 5.5 | $0.00006 | $0.00253 |
| Haiku 4.5 | $0.00003 | $0.00126 |
Grade A, and why
verify-upstream-config-examples scanned grade A with 1 finding 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 1mo 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.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
- **claude**: `curl -sL https://code.claude.com/llms.txt` — produces a flat list of doc URLs. Grep for lines containing `hooks` or `mcp`; curl each linked page (also markdown). How it starts
The opening of the file, as written. The whole thing — 129 lines — stays where its author put it; the contents beside it link to each section on GitHub.
verify-upstream-config-examples
Detects drift between the vendored corpus under testdata/upstream-examples/ and live vendor docs. Net-new examples that pass testagent <vendor> validate --strict are written to the corpus and surfaced as a draft PR.
Vendor dispatch
| Phrase | Vendor arg |
|---|---|
| "verify upstream config examples" (no vendor named) | ask the user or default to all |
| "verify claude config corpus" / "check claude drift" | claude |
| "verify codex config corpus" / "check codex drift" | codex |
| "verify cursor config corpus" / "check cursor drift" | cursor |
| "verify upstream config examples all" | all (runs all three sequentially) |
Workflow
1. Scope check
Vendor must be claude, codex, cursor, or all. If ambiguous, ask: "Which vendor — claude, codex, cursor, or all?"
2. Fetch the vendor's source
Per sources.md:
- claude:
curl -sL https://code.claude.com/llms.txt— produces a flat list of doc URLs. Grep for lines containinghooksormcp; curl each linked page (also markdown). - cursor:
curl -sL https://cursor.com/llms.txt— same shape. Linked pages under/docs/hooksand/docs/context/mcp. - codex:
gh api repos/openai/codex/contents/README.md -H 'Accept: application/vnd.github.raw'— inline fenced blocks; no link-following needed.
3. Extract fenced code blocks
From each fetched page, extract all fenced code blocks of the relevant language:
claudeandcursor:```jsonblocks.codex:```tomlblocks.
Use awk to detect ```json / ```toml openers and matching ``` closers. Produce one candidate file per block.
4. Diff against local corpus
For each candidate block:
- Compute SHA256 of the block content.
- Compare against SHA256 of every file in
testdata/upstream-examples/<vendor>/. - Skip blocks already vendored verbatim (exact content match).
- Flag blocks where a same-name fixture exists but content differs (drift in an existing fixture).
What ships with it
1 file 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.
- 1mo ago First seen · 129 lines · 29 tokens per session scan A 46bd9b6c327b
verify-upstream-config-examples is a skill published in the GitHub repository paultyng/testagent (7 stars, last pushed 1mo ago), licensed MIT. It adds 29 tokens to every session and 1,263 once invoked, about $0.0001 per session on Opus 5.5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
verification-loop
Evidence-before-assertions workflow. Use before claiming work is done, before release, and after any behavior change in scripts/skills/MCP.
vision-verifier
Use when verifying Computer Vision imports, fixtures, models, schemas, cameras, artifacts, cleanup, latency, and stated limitations with reproducible evidence.
orchestrating-adversarial-reviews
Multi-agent adversarial-verification orchestration for high-confidence conclusions. Fan-out finders, then verify every finding through a three-prism panel (exploitability / correctness / refutation) that defaults to disbelief, gate fixes behind load-bearing proof tests that catch agents who falsely claim "done/fixed"…
aitask-qa
Run QA analysis on any task — analyze changes, discover test gaps, run tests, and create follow-up test tasks.
aitask-resume
Resume an in-flight task from its gate-ledger checkpoint (programmatic re-entry surface for testing / TUI / board ops).
aitask-audit-wrappers
Audit and port aitask skill wrappers across code-agent trees, plus helper-script whitelist coverage.