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 stefanobaghino/simple-output-styles --skill feedbackgit clone --depth 1 https://github.com/stefanobaghino/simple-output-stylesWrote 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/stefanobaghino/simple-output-styles/feedback)<a href="https://agentmods.dev/skills/stefanobaghino/simple-output-styles/feedback"><img src="https://agentmods.dev/badge/skills/stefanobaghino/simple-output-styles/feedback.svg" alt="Measured on agentmods" 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 | $0.00072 | $0.00716 |
| Opus 5 | $0.00036 | $0.00358 |
| Sonnet 5 | $0.00014 | $0.00143 |
| Haiku 4.5 | $0.00007 | $0.00072 |
Grade B, and why
feedback scanned grade B 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 8d 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.
Reads agent configuration directoriesmediumAgent snooping
.claude/, .codex/, .gemini/ hold keys, settings and other credentials a mod has no legitimate need for.
then `~/.claude/settings.json`. Confirm it with the user, since the How it starts
The opening of the file, as written. The whole thing — 64 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Provide feedback on simple-output-styles
Help the user open the right issue on the simple-output-styles repository with as little typing as possible. You gather the facts; the user reviews a prefilled form and submits it.
Do not open the issue for the user with gh or the API. The repository closes
issues that bypass the forms. Always hand the user a prefilled form link.
Step 1: pick the form
Ask the user one question: is the feedback about an existing style (actionable-clarity, clarity-flow, classic-concise, developer-docs, plain-language, technical-simplified), or about something else (evals and methodology, documentation, installation, licensing and attribution, a new style request, or anything else)? Skip the question if the request already makes the answer clear.
Step 2a: style feedback
Gather each field, asking the user only for what you cannot find yourself:
- Version: run
claude --versionand use the output. - Styles: find the active output style: check
outputStylein.claude/settings.local.jsonand.claude/settings.jsonin the project, then~/.claude/settings.json. Confirm it with the user, since the feedback may be about a style other than the active one. If you find nothing and the user does not know, useI don't know. - Sentiment: ask whether the feedback is overall positive, mixed, or negative, unless the conversation already makes it obvious.
- Feedback: ask for a short description if the user has not given one. If the feedback is about output produced in this conversation, offer to include the relevant excerpt.
Build the link (URL-encode every value; join multiple styles with commas):
https://github.com/stefanobaghino/simple-output-styles/issues/new?template=style-feedback.yml&styles=<styles>&version=<version>&sentiment=<Positive|Mixed|Negative>&feedback=<text>
Step 2b: general feedback
Ask for the area if it is not clear — one of Evals and methodology,
Documentation, Installation, New style request,
Licensing and attribution, Other — plus the sentiment and a short
description, then build the link (URL-encode every value):
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.
- 8d ago First seen · 64 lines · 72 tokens per session scan B d72be4498214
feedback is a skill published in the GitHub repository stefanobaghino/simple-output-styles (16 stars, last pushed 3d ago), licensed 0BSD. It adds 72 tokens to every session and 716 once invoked, about $0.0004 per session on Opus 5. A static security scan graded it B with 1 finding (reads agent configuration directories). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
amend-guide
The requirements changed while someone is halfway through a learn-as-you-go guide — fold the change in WITHOUT rewriting the work they've already executed. Reads foundation/progress.md to find the frontier (and stops to have it confirmed if it's missing or stale), verifies any new fact online, then delivers an impact…
check-my-work
Reconcile the REAL project against the learn-as-you-go guide you have been following — does what you built actually contain what the steps you executed told you to write? Reads the milestone checkpoints (the complete-file sections of each NNverify.md), diffs them against your files while ignoring cosmetic differences…
mark-progress
Record what you have actually EXECUTED while following a learn-as-you-go guide — step by step — into the guide's foundation/progress.md ledger, and reconcile foundation/status.md (frontier, milestone table, session log) with it. Marks only: it never edits step files, never fixes anything, and never marks a milestone ✅…
report-issue
A reader hit a real issue while following a learn-as-you-go guide — report it and fix the guide so the issue can't recur. Diagnoses the root cause, fixes the step where they got stuck, then sweeps the WHOLE guide for every other place the same class of defect appears and fixes those too, adds a failure-note guard…
draft-milestone
Draft an approved learn-as-you-go build guide into folders of atomic, teaching step files — the WHOLE guide by default (every milestone, in one pass), or a single named milestone if one is given (for re-drafts/fixes). Use after a plan from the plan-guide skill has been approved. Invoke with no argument to draft the…
log-feedback
A reader hit friction while following a learn-as-you-go guide — capture it to the guide's feedback log WITHOUT changing the guide. Appends one dated, structured entry (where · reader · what happened · suspected class · severity · tags · quote) to guide/feedback-log.md, so friction becomes durable data for improving…