write-skill

A guide for creating or improving coding-agent skills, including their trigger descriptions, instructions, supporting references, and tests.

In plain words
What is it for?
Use it to write a new SKILL.md, fix a skill that triggers too often or too rarely, or tune one using benchmark or session results.
Why use it?
It helps make a skill activate for the right requests and change the agent’s behavior without adding unnecessary instructions.

Skill for Claude CodeCodex

Install

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.

agentmods
npx agentmods add skills/baleen37/bstack/write-skill
Any agent
npx skills add baleen37/bstack --skill write-skill
Clone the repo
git clone --depth 1 https://github.com/baleen37/bstack

Made for: Claude Code, Codex.

Per session 52 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 959 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5 $0.00052 $0.00959
Opus 5 $0.00026 $0.00479
Sonnet 5 $0.00010 $0.00192
Haiku 4.5 $0.00005 $0.00096

Measured yesterday against content hash c74a9d54847a, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

write-skill 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 yesterday.

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.

plugins/me/skills/write-skill/SKILL.md · 95 lines

How it starts

The opening of the file, as written. The whole thing — 95 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Write Skills

Write the smallest SKILL.md that changes behavior. Most skills are one file with two frontmatter fields and a short body.

---
name: my-skill            # lowercase-hyphens, matches the directory name
description: Use when ... # third person, names the trigger, <=1024 chars
---

The description is the whole trigger mechanism — it is all the model sees before deciding to load the skill. Say when to use it, using the words someone would actually type, and cover both failure directions ("triggers too rarely or too often", not just one). Never write "I help you..." .

For the body: include only what the model gets wrong without you. It knows the domain; it does not know your conventions or the traps. Cut anything else — a skill that restates common knowledge changes no behavior and costs context every session. Keep it under 500 lines, and write rules that hold for the whole task rather than one-time steps, because the body is injected once and never re-read.

Split to references/ only what a subset of runs needs, link it one level deep from here, and say what is in it so the agent knows whether to open it.

Prove it works

Skip this and you are guessing. A new skill's baseline is the same task with no skill at all.

  1. Write 3+ tasks that show the failure, with a checkable success signal.
  2. Run them with no skill. That is the baseline.
  3. Write the smallest thing that fixes it.
  4. Re-run. Keep it only if it beats baseline.

If baseline already passes, do not write the general skill — say so. Then look for what the model cannot derive: your environment's facts, your conventions, the traps it knows in theory but misses under pressure. Either narrow the skill to that and re-test, or report that no skill is warranted. A skill that restates what baseline already did is pure context cost.

Match the testing to the stakes. A one-screen skill deserves a trigger check and one real task, not seven subagents.

Test the description separately: give a subagent the frontmatter only, with 5 prompts that should trigger and 4 that should not.

Read the full file on GitHub · 95 lines

Files

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.

Changes

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.

  1. yesterday First seen · 95 lines · 52 tokens per session scan A c74a9d54847a

Subscribe to this mod's changes

write-skill is a skill published in the GitHub repository baleen37/bstack (4 stars, last pushed 11d ago), licensed MIT. It adds 52 tokens to every session and 959 once invoked, about $0.0003 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.

Related

Other skills, from other repositories

systematic-debugging

Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.

obra/superpowers · 21 tokens

next-cache-components-adoption

Turn on Cache Components in a Next.js app and resolve the blocking routes it surfaces. Use when the user wants to enable, adopt, or migrate to Cache Components, flip the cacheComponents flag, work through a flood of blocking-prerender / instant validation errors, run the cache-components-instant-false codemod, or…

vercel/next.js · 95 tokens

babysit-pr

Babysit a GitHub pull request after creation by continuously polling review comments, CI checks/workflow runs, and mergeability state until the PR is merged/closed or user help is required. Diagnose failures, retry likely flaky failures up to 3 times, auto-fix/push branch-related issues when appropriate, and keep…

openai/codex · 114 tokens

imagegen

Generate or edit raster images when the task benefits from AI-created bitmap visuals such as photos, illustrations, textures, sprites, mockups, or transparent-background cutouts. Use when Codex should create a brand-new image, transform an existing image, or derive visual variants from references, and the output…

openai/codex · 113 tokens

cpu-profile-analysis

Analyze V8/Chrome CPU profiles (.cpuprofile) and DevTools trace files (Trace-.json). Use when: profiling performance, investigating slow functions, comparing code paths, finding bottlenecks, analyzing timeToRequest, understanding call trees from sampling profiler data, analyzing layout/paint/rendering, investigating…

microsoft/vscode · 71 tokens

next-cache-components-optimizer

Drive a Next.js route to instant navigation by setting up an agentic loop, under Cache Components / PPR, on initial load (hard navigation) and client-side navigation (soft navigation). Encode the goal as a failing @next/playwright instant() e2e and work it to green, one verified route at a time; the shipped test then…

vercel/next.js · 170 tokens