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 pe-menezes/vibeflow --skill vibeflow-hotfixgit clone --depth 1 https://github.com/pe-menezes/vibeflowWrote 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/pe-menezes/vibeflow/vibeflow-hotfix)<a href="https://agentmods.dev/skills/pe-menezes/vibeflow/vibeflow-hotfix"><img src="https://agentmods.dev/badge/skills/pe-menezes/vibeflow/vibeflow-hotfix/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/pe-menezes/vibeflow/vibeflow-hotfix"><img src="https://agentmods.dev/badge/skills/pe-menezes/vibeflow/vibeflow-hotfix.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 | $0.00103 | $0.02378 |
| Opus 5 | $0.00051 | $0.01189 |
| Sonnet 5 | $0.00021 | $0.00476 |
| Haiku 4.5 | $0.00010 | $0.00238 |
Grade A, and why
vibeflow-hotfix 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 10d 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.
This is a copy
86% identical to hotfix — 63 lines differ, which has more behind it and is treated as the original. This page carries a canonical link to it rather than competing with it.
How it starts
The opening of the file, as written. The whole thing — 214 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Vibeflow: Hotfix
In one call, produces a short trace doc in .vibeflow/hotfixes/ and
implements the fix, born with a regression test that failed before it.
Exactly two outcomes: done or a named HALT that saves the doc and hands the
next step to a human. Done carries the doc, the fix, and the regression
test — red→green when the host runs tests; when it cannot, the doc declares
verification: not-run and the status locks at partial. Production
incidents are the motivating case, not a gate. The discriminator against
the other skills is evidence — an observed defect you can show, wherever
it surfaced.
Usage: Provide the symptom + evidence as input, or the path to a halted doc to resume it.
Language
Detect the language of the user's input. Write ALL output in that same language. Technical terms in English are acceptable regardless of the detected language. One exception: the trace doc's section names, field names, and vocabulary values stay literally in English in any output language — resuming and consolidation match them exactly.
The one-call contract
You are a coding agent on a live defect: the evidence decides, you execute
and record. One invocation, exactly two structured outcomes — done or a
named HALT. No mandatory questions in between: a question you would need
answered becomes the HALT that motivates it, with the doc saved and the
question recorded as the human's next step. Degradation is declared, not
silent — a host that cannot run tests writes verification: not-run in the
doc and the status locks at partial.
The trace doc
.vibeflow/hotfixes/<YYYY-MM-DD>-<slug>.md — one file per bug; create the
directory if it doesn't exist. Written in two moments: Time 1 before any code
edit, Time 2 after the fix. Two rules make it auditable:
Symptomis immutable once written. It exists before the fix, so the fix cannot rationalize it after the fact.Eliminated / EvidenceandDeviationsare append-only.
# Hotfix: <slug>
origin: session | prompt | third-party | resumed
status: verified | partial | halted(<condition>)
## Symptom
<the observed behavior, with the real data — immutable once written>
## Checkpoint
hypothesis: <current explanation — overwrite when evidence changes it>
falsification_test: <the observation that would prove the hypothesis wrong>
blind_spots: <what the hypothesis does not explain, or was not checked>
## Preservation
- <1–3 properties the system SHALL CONTINUE TO satisfy after the fix>
## Eliminated / Evidence
<materializes only when hypothesis 1 fails; append: hypothesis — the evidence that killed it>
## Root cause
<the defect, located>
## Fix
files_changed: <paths>
<what changed and why it closes the cause>
## DoD
- [ ] <2–4 binary checks>
## Regression
WHEN <trigger, with the data> THEN <expected behavior>
test: <path of the test in the project's suite — written at the red proof>
oracle_type: specified | derived | metamorphic | implicit
reproduction: real | synthetic | none
verification: red-green | not-run
## Deviations
<append-only: what diverged from the assessment, and why>
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.
- 10d ago First seen · 214 lines · 103 tokens per session scan A 83503a804b06
vibeflow-hotfix is a skill published in the GitHub repository pe-menezes/vibeflow (29 stars, last pushed 4d ago), licensed MIT. It adds 103 tokens to every session and 2,378 once invoked, about $0.0005 per session on Opus 5. A static security scan graded it A with 0 findings. It is 86% identical to hotfix, differing in 63 lines, and is treated as a copy.
Other skills, from other repositories
qa-knowledge
To run QA engineering — requirements/gap analysis, scenario & spec design, test implementation, failure triage — over the QA knowledge base.
megalinter-check
Collect MegaLinter lint errors for the current repository. Use when the user wants to know if the code passes linting, why the MegaLinter CI job fails, or before/after fixing lint errors. Two modes - watch a CI job (GitHub Actions, GitLab CI, Azure Pipelines, Bitbucket Pipelines) and parse its logs, or run MegaLinter…
anchor-repro
Reproduce a behavioral bug before fixing it, record the failing probe, and verify the fix with the same probe. Use for bug reports, failing tests, crashes, hangs, regressions, wrong output, and observable behavior that should change.
reqforge-greenkeeper
name: reqforge-greenkeeper description: Used when maintaining the ReqForge repository itself and its release gates fail — pnpm test, pnpm forge-smoke, pnpm sync:discover, adapter drift, skill fixtures, loadouts, or test-demo golden path. Restores a green, synced state with minimal changes. Not for bugs in user apps…
analyze-failures
Triage Katalon True Platform/TestOps test failures and file defects. Use when you need to investigate failed test results, classify each failure as product defect vs automation defect vs environment/data issue, cluster failures by common signature, find likely root cause from execution data, and optionally create…
megalinter-fix
Fix the errors reported by MegaLinter. Use after megalinter-check found errors, or when the user pastes MegaLinter/CI lint errors and wants them fixed. Applies safe fixes automatically (auto-fix linters first, then guided manual fixes using per-linter fix guides), asks the user when fixing is ambiguous, and can…