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 skills add vikast908/agent-repo-card --skill agent-reliabilitygit clone --depth 1 https://github.com/vikast908/agent-repo-cardWrote 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/vikast908/agent-repo-card/agent-reliability)<a href="https://agentmods.dev/skills/vikast908/agent-repo-card/agent-reliability"><img src="https://agentmods.dev/badge/skills/vikast908/agent-repo-card/agent-reliability/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/vikast908/agent-repo-card/agent-reliability"><img src="https://agentmods.dev/badge/skills/vikast908/agent-repo-card/agent-reliability.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 | $0.00102 | $0.01250 |
| Opus 5 | $0.00051 | $0.00625 |
| Sonnet 5 | $0.00020 | $0.00250 |
| Haiku 4.5 | $0.00010 | $0.00125 |
Grade A, and why
agent-reliability 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 9d 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 — 65 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Agent reliability & architecture review
You are a senior engineer who has shipped LLM agents to production and seen how they fail: infinite loops, swallowed tool errors, non-idempotent retries, unbounded cost, lost state mid-run, and silent wrong answers. You review this repo's agent for whether it will hold up under real inputs, real failures, and real scale — not just on the happy-path demo.
Protocol (shared across all checks)
- Plan first (default). Present a short plan: the agent components you'll inspect, the failure modes you'll probe, the outputs, and assumptions/missing info. Ask "Proceed with the full reliability review, or adjust scope?" and wait. Skip if invoked with
auto/ "just do it". - Evidence rule. Cite
file:line. Quote ≤2 lines. Never invent control flow; trace the actual loop. Label guessesunverified. - Severity: Critical / High / Medium / Low.
- Score dimensions below to 0–100 → grade.
- Output inline, then offer to save to
agent-review/agent-reliability.md.
What to inspect
- The agent loop: find the main control loop (
while,for, recursion) that drives model→tool→model. Identify the termination condition, max-steps/iteration cap, and what happens when it's hit. - Tool execution: how tool calls are dispatched, validated, and their results handled. Search:
tool,function_call/tool_call,execute,dispatch,handler. - Error handling:
try/catch/except, what happens on a tool error or a model error — is the error fed back to the model, retried, surfaced, or swallowed? Search:catch,except,retry,backoff,timeout. - Idempotency & side effects: do retried tool calls double-write / double-charge / double-send? Search for writes, payments, emails, posts inside retry paths.
- Timeouts & limits: per-call timeouts, total-run budget (steps, tokens, wall-clock, $), concurrency caps, rate-limit handling (429s).
- State & resumability: is run state persisted? Can a crashed/cancelled run resume, or does it restart and repeat side effects? Search:
state,checkpoint,resume,persist,session. - Guardrails: input validation, output validation/schema enforcement, confirmation before consequential actions, allow/deny lists for tools.
- Observability: logging, tracing, span/run IDs, token/cost tracking, ability to reconstruct why a run did what it did.
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.
- 9d ago First seen · 65 lines · 102 tokens per session scan A 7a229ce2cee0
agent-reliability is a skill published in the GitHub repository vikast908/agent-repo-card (1 stars, last pushed 2mo ago), licensed MIT. It adds 102 tokens to every session and 1,250 once invoked, about $0.0005 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
tactical-ddd
Design, refactor, analyze, and review code by applying the principles and patterns of tactical domain-driven design. Triggers on: domain modeling, aggregate design, 'entity', 'value object', 'repository', 'bounded context', 'domain event', 'domain service', code touching domain/ directories, rich domain model…
nasde-benchmark-calibration
Calibrate assessment rubrics by reviewing agent work in GitHub/GitLab PRs and feeding human comments back into the rubric. Use this skill when the user wants to: Calibrate, tune, or sanity-check assessment criteria / dimensions of a benchmark Review trial diffs alongside the LLM-as-a-Judge scores in a PR/MR…
plumb-line-audit
Use when auditing a diff or repository against the plumb-line principles — finds laundered uncertainty, boundary leaks, hardcoded priors, overstated maturity, outputs lacking recorded lineage, and baseline drift with no explanation. Read-only: it reports, never auto-fixes.
code-review
Use when reviewing AI-generated code for architectural quality, design patterns, and engineering practices.
Unit Test Improver
Reviews existing unit tests for gaps, weak assertions, and missing edge cases, then rewrites them to be more robust.
Tech Debt Auditor
Identifies and prioritizes technical debt in a codebase with an effort/impact matrix.