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 calionauta/stelow --skill stelow-workflow-testing-ai-codegit clone --depth 1 https://github.com/calionauta/stelowWrote 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/calionauta/stelow/stelow-workflow-testing-ai-code)<a href="https://agentmods.dev/skills/calionauta/stelow/stelow-workflow-testing-ai-code"><img src="https://agentmods.dev/badge/skills/calionauta/stelow/stelow-workflow-testing-ai-code/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/calionauta/stelow/stelow-workflow-testing-ai-code"><img src="https://agentmods.dev/badge/skills/calionauta/stelow/stelow-workflow-testing-ai-code.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 3 findings, up to medium
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- medium MCP Rug Pull · line 351 npx commands without a version suffix (e.g. @1.0.0) create a rug-pull risk if the upstream server is compromised and publishes a malicious update.Fix: Pin the version: npx @scope/[email protected]
- medium MCP Rug Pull · line 434 npx commands without a version suffix (e.g. @1.0.0) create a rug-pull risk if the upstream server is compromised and publishes a malicious update.Fix: Pin the version: npx @scope/[email protected]
- medium MCP Rug Pull · line 437 npx commands without a version suffix (e.g. @1.0.0) create a rug-pull risk if the upstream server is compromised and publishes a malicious update.Fix: Pin the version: npx @scope/[email protected]
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 | $0.00104 | $0.03324 |
| Opus 5.5 | $0.00042 | $0.01330 |
| Sonnet 5.5 | $0.00021 | $0.00665 |
| Haiku 4.5 | $0.00010 | $0.00332 |
Grade A, and why
stelow-workflow-testing-ai-code 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 14d 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 — 344 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AI-Aware Testing Strategy
Based on empirical research:
- AgentAssay (2026): Non-deterministic agent testing framework
- MSR 2026: Over-mocking anti-patterns in AI-generated tests
- Veracode 2025: 45% of AI code contains vulnerabilities
- CodeRabbit 2025: AI code has 1.7x more bugs than human code
- CoderEval (2023): 43.1% of AI code is less robust
- SWE-Mutation, ACL Findings 2026: LLM suites are superficial; semantic mutants drop detection 71% → 40%
- VibeCheck, Sep 2026: execution-adequacy gap — runnable tests lack strong assertions and edge coverage
- Hamidi et al., Sep 2026: oracles fail on the faults that matter; assertions need manual reasoning
- TDAD 2026: graph test-context beats procedural TDD instructions (−70% regressions); concise skills win
- TDFlow, EACL 2026: human-written tests 94.3% vs self-generated 68% — test authorship is the hurdle
- Kun Chen ProgramBench (Jun 2026): mandated TDD skill hurt pass rate (−3.6pp, p<0.0001) at +55% cost — self-written tests lock in incomplete scope, green reads as done.
- Kun Chen DeepSWE (Oct 2026): agent-written unit (65%) + integration (35%) tests added zero success-rate gain vs writing none; executing existing suites in-loop also measured ~zero. Human-described cases may still help (unproven); e2e inconclusive.
Standalone awareness: when inside stelow, triggered automatically by product_type in spec-product frontmatter. When standalone, invoke directly with a spec-product path (spec-product*.md). Appetite defaults to Core if not found — documented in output. All test-breadth tables and quality baselines work identically in both modes.
Activation
- Trigger:
product_type: softwareorproduct_type: hybridin spec-product frontmatter - Phase: Phase 11 (Tech Planning)
- Prerequisite: approved spec-product with scope defined
Step 2: Read Appetite and Product Context
Read quality from the spec-product file (spec-product*.md) before generating test scopes. When running standalone, quality defaults to production if not found in frontmatter — the skill documents this assumption in the output. A legacy appetite: line is accepted and mapped once (any legacy value resolves to production; breadth 2/3/5 from Lean/Core/Complete).
What ships with it
3 files 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.
- 14d ago Changed · -3 lines b5a19eb0709c
- 28d ago Changed · +8 lines 24baca9ca871
- 1mo ago Changed · +1 lines 2530c783a64c
- 1mo ago Changed · -1 lines d4e59e7ab5d1
- 1mo ago Changed · +6 lines acefb2a88b2b
- 1mo ago Changed · +14 lines 46ef2e01dacc
- 1mo ago Changed · +48 lines e54f70779fbd
- 1mo ago Changed · +5 lines b6ea0e7b33a5
- 1mo ago Changed · -366 lines a9c70877e45d
- 2mo ago First seen · 632 lines · 104 tokens per session scan A 40396392eb2e
stelow-workflow-testing-ai-code is a skill published in the GitHub repository calionauta/stelow (14 stars, last pushed 9d ago), licensed MIT. It adds 104 tokens to every session and 3,324 once invoked, about $0.0004 per session on Opus 5.5. 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.
stelow-workflow-testing-ai-code compared with skills from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| stelow-workflow-testing-ai-codecalionauta/stelow | Skill | 14 | 9d ago | MIT | A | 104 | Claude Code, Codex |
| imagegenopenai/codex | Skill | 133k | yesterday | Apache-2.0 | A | 113 | Codex |
| playwright-climicrosoft/playwright | Skill | 99k | yesterday | Apache-2.0 | A | 19 | Claude Code |
| local-ai-agentsmicrosoft/ai-agents-for-beginners | Skill | 79k | 4d ago | MIT | A | 200 | Claude Code, Codex |
| systematic-debuggingobra/superpowers | Skill | 313k | 3d ago | MIT | A | 21 | Claude Code |
| next-cache-components-adoptionvercel/next.js | Skill | 143k | yesterday | MIT | A | 95 | Claude Code |
| next-partial-prefetching-adoptionvercel/next.js | Skill | 143k | yesterday | MIT | A | 103 | Claude Code |