agent/agentskills
2
0
Fork 0
mirror of https://github.com/agentskills/agentskills.git synced 2026-06-18 15:54:06 +08:00
Code Issues Packages Projects Releases Wiki Activity
101 Commits 9 Branches 0 Tags
4ae320a9f205c0dc4bbb65342d500e7ed3cb12f0
Commit Graph

83 Commits

  • Add nanobot to Agent Skills client showcase
    chengyongru · 2026-04-14 10:45:14 +08:00
    4cc096a7ab
  • Add source code url to Google AI Edge Gallery in client list
    Jing Jin · 2026-04-10 13:40:02 -07:00
    f4f20dc8c5
  • Scale down Google AI Edge Gallery logo in clients.jsx 
    The logo is a 256x256 square icon with no wordmark, so at default scale
it appears much larger than the horizontal wordmark logos used by other
clients. Setting `scale: 0.45` brings it in line visually.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
    Jonathan Hefner · 2026-04-10 09:35:47 -05:00
    acaa82180c
  • Merge pull request #307 from jinjingforever/add_gallery 
    Add Google AI Edge Gallery to Agent Skills client list
    Jonathan Hefner · 2026-04-10 09:24:38 -05:00
    650984699b
  • Merge pull request #283 from jonathanhefner/allowed-tools-string 
    Use precise type name for `allowed-tools` field
    Jonathan Hefner · 2026-04-10 09:21:31 -05:00
    1eeb1aab05
  • Add Google AI Edge Gallery to Agent Skills client list
    Jing Jin · 2026-04-09 16:21:09 -07:00
    60830f1884
  • Improve client descriptions and names in clients.jsx 
    Update descriptions to use better-curated quotes from vendor websites.
Also make display names more specific where the entry refers to a
product rather than the parent company (e.g. "GitHub" → "GitHub
Copilot", "Databricks" → "Databricks Genie Code", "Snowflake" →
"Snowflake Cortex Code").

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
    Jonathan Hefner · 2026-04-02 09:46:51 -05:00
    cbc20299ae
  • Add shuffle/alphabetical toggle to ClientShowcase 
    Clients are now shuffled by default so that alphabetical ordering
doesn't unfairly prioritize certain clients. A segmented icon toggle
(shuffle / A-Z) in the top-right corner lets users switch between
shuffled and alphabetical order. Each click of the shuffle button
produces a fresh random order.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
    Jonathan Hefner · 2026-04-01 17:39:12 -05:00
    ee61ae3cf9
  • Add Client Showcase page 
    Add a dedicated page listing all clients that support the Agent Skills
format. The `ClientShowcase` component renders the shared `clients` data
as a 2-column grid of cards, sorted alphabetically, with each card
showing the client logo, description, and links to setup instructions
and source code when available.

Add `"clients"` to the navigation in `docs.json`.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
    Jonathan Hefner · 2026-04-01 17:38:42 -05:00
    6f18cf6609
  • Extract client data from LogoCarousel into shared clients.jsx 
    Move the hardcoded client data array out of `LogoCarousel.jsx` into a
separate `docs/snippets/clients.jsx` module so it can be reused by other
components. `LogoCarousel` now accepts a `clients` prop, and `home.mdx`
imports and passes the shared data. Also extract a `Logo` sub-component
to reduce duplication of the light/dark `<img>` pair.

Update `CONTRIBUTING.md` to point contributors to `clients.jsx`.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
    Jonathan Hefner · 2026-04-01 15:15:26 -05:00
    b10b93a438
  • Override Mintlify's default <img> margins in logo carousel 
    Mintlify adds 36px top and bottom margins to `<img>` elements in the
main content area. This caused unwanted vertical spacing in the logo
carousel rows. Adding `!my-0` to each `<img>` removes those margins.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
    Jonathan Hefner · 2026-04-01 15:08:19 -05:00
    ee7d80cb8b
  • Move logo sizing from <img> to wrapper <div> in logo carousel 
    Use a wrapper `<div>` to control each logo's width instead of styling
the `<img>` elements directly. This also replaces inline styles on the
`<a>` elements with Tailwind utility classes (`no-underline`,
`border-none`) and simplifies the map callbacks.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
    Jonathan Hefner · 2026-04-01 15:08:19 -05:00
    15ddb74928
  • Replace per-logo width with relative scale in logo carousel 
    The logo carousel previously used a per-logo `width` property (e.g.
`"120px"`, `"200px"`) with a fallback default of `"150px"`. This made
logo sizing an absolute, context-specific value tied to the carousel
layout.

Replace `width` with a relative `scale` multiplier (e.g. `0.8`, `1.33`)
that each consuming component can apply to its own base width. This
prepares for sharing the logo data array between `LogoCarousel` and
other components, where the base width might differ. Logos without a
`scale` property default to 1.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
    Jonathan Hefner · 2026-04-01 09:35:16 -05:00
    8c2596540b
  • Add description field to each logo carousel entry 
    Vendor descriptions sourced verbatim from each client's own website or
GitHub repository. The `description` field is placed immediately after
`name` in each entry, grouping identity metadata together.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
    Jonathan Hefner · 2026-03-30 16:19:07 -05:00
    15065098a2
  • Update OpenHands URL from all-hands.dev to openhands.dev 
    The old URL `https://www.all-hands.dev/` now redirects to
`https://openhands.dev/`.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
    Jonathan Hefner · 2026-03-30 16:02:52 -05:00
    028a8fef9b
  • Use precise type name for allowed-tools field 
    Replace "space-delimited list" with "space-separated string" in both the
frontmatter table and the `allowed-tools` field section. This better
reflects that the value is a YAML scalar, per the given example.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
    Jonathan Hefner · 2026-03-30 12:11:35 -05:00
    6f92fcdb78
  • Merge pull request #279 from jonathanhefner/fix-carousel-logo-takeover 
    Add `noZoom` to logo carousel images
    Jonathan Hefner · 2026-03-27 12:12:41 -05:00
    5e7f3e2c4c
  • Add noZoom to logo carousel images 
    Mintlify enables click-to-zoom on all `<img>` elements by default. This
caused clicking a carousel logo to both open the vendor link in a new
tab AND zoom the logo to fullscreen in the original tab. Adding `noZoom`
disables the lightbox so only the link navigation fires.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
    Jonathan Hefner · 2026-03-27 10:05:14 -05:00
    5343c3f42b
  • Replace nested hexagon favicon with solid hexagon 
    Remove the inner hexagon path with white stroke from `docs/favicon.svg`,
leaving a single solid black hexagon to match the updated logo.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
    Jonathan Hefner · 2026-03-27 09:30:13 -05:00
    85ef34c715
  • Replace shuf with $RANDOM in quickstart for broader compatibility 
    `shuf` is a GNU coreutil which requires `brew install coreutils` on
macOS. Replace it with `echo $((RANDOM % <sides> + 1))` which works in
bash and zsh on both Linux and macOS.

Also reword "true randomness" to "using a random number generator" to
more clearly distinguish programmatic RNG from LLM non-determinism.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
    Jonathan Hefner · 2026-03-23 11:25:26 -05:00
    d7dba49441
  • Merge pull request #274 from jonathanhefner/discord-link 
    Add Discord links to `README.md` and `docs/home.mdx`
    Jonathan Hefner · 2026-03-23 10:44:50 -05:00
    d91d26d0ab
  • Add Discord links to README.md and docs/home.mdx 
    Adds a Discord entry to the Getting Started list in `README.md` and
replaces the standalone "View on GitHub" link in `docs/home.mdx` with a
sentence inviting readers to join on GitHub or Discord.

Also applies minor formatting cleanup in `README.md`: bolds the link
text and switches hyphens to em dashes for consistency.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
    Jonathan Hefner · 2026-03-23 10:41:25 -05:00
    84672e62b9
  • Merge pull request #261 from nairteashop/add-kiro-logo 
    Add Kiro logo to carousel.
    Jonathan Hefner · 2026-03-20 10:40:12 -05:00
    2fbfc322d5
  • Merge pull request #253 from jonathanhefner/compatibility-field-show-versioning 
    Add runtime version example to `compatibility` field
    Jonathan Hefner · 2026-03-19 18:24:03 -05:00
    5bb99345b3
  • Add Kiro logo.
    Arun Nair · 2026-03-19 16:08:01 -07:00
    cc89b9c41a
  • Add runtime version example to compatibility field 
    Add a third example showing a Python version requirement alongside `uv`.
This complements the existing examples (product targeting and system
tool requirements) by demonstrating version-pinned runtime requirements
— the kind of non-obvious constraint that the `compatibility` field is
useful for.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
    Jonathan Hefner · 2026-03-16 12:05:26 -05:00
    99b8edf685
  • Add quickstart guide for skill creators 
    Adds a tutorial-style quickstart guide that walks the reader through
creating their first Agent Skill — a `roll-dice` skill that teaches an
agent to roll dice using true system randomness. The guide covers
creating the `SKILL.md` file, verifying discovery via `/skills` in VS
Code, testing with a "Roll a d20" prompt, and a brief explanation of the
discovery/activation/execution lifecycle. Includes both bash and
PowerShell command variants, and a note about model variation in
tool-use reliability.

Also adds the quickstart as the first page under "For skill creators" in
`docs.json` navigation.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
    Jonathan Hefner · 2026-03-16 10:35:35 -05:00
    3fccd701c1
  • Merge pull request #248 from jonathanhefner/add-gotchas-section-to-best-practices 
    Add gotchas section to best practices guide
    Jonathan Hefner · 2026-03-16 09:50:36 -05:00
    e42cbd6f8a
  • Merge pull request #240 from jonathanhefner/clarify-skill-eval-file-tree 
    Add skill directory to structure diagram in evaluating skills doc
    Jonathan Hefner · 2026-03-16 09:47:42 -05:00
    5cd19050c9
  • Add gotchas section to best practices guide 
    Add a new "Gotchas sections" subsection to "Patterns for effective
instructions" in the best practices guide. Covers what a gotcha is
(environment-specific facts that defy reasonable assumptions), how to
structure them (problem/correction pairs), and why they belong directly
in `SKILL.md` rather than a separate reference file.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
    Jonathan Hefner · 2026-03-15 17:32:38 -05:00
    4be62ebad0
  • Add skill directory to structure diagram in evaluating skills doc 
    Show `csv-analyzer/` (containing `SKILL.md` and `evals/evals.json`)
alongside `csv-analyzer-workspace/` so readers can see the full layout
at a glance.

Closes #238. Closes #239.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
    Jonathan Hefner · 2026-03-13 10:59:58 -05:00
    7d9e51a92a
  • Add best practices guide for skill creators 
    Covers grounding skills in real expertise (extracting from tasks,
synthesizing from project artifacts), iterating with execution feedback,
managing context budget, calibrating instruction specificity, and
reusable instruction patterns (templates, checklists, validation loops,
plan-validate-execute, script bundling).

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
    Jonathan Hefner · 2026-03-10 14:19:12 -05:00
    db92534bfb
  • Remove redundant intro sentence from specification doc 
    The frontmatter `description` already serves as the page subtitle in
Mintlify, so the "This document defines the Agent Skills format." line
was just restating it.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
    Jonathan Hefner · 2026-03-10 14:16:18 -05:00
    8c82ac2a8d
  • Add ... entry to directory tree in specification doc 
    The directory listing is not exhaustive — skill authors can include
other files and subdirectories beyond those shown. Adding a `...` entry
makes this explicit.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
    Jonathan Hefner · 2026-03-10 14:16:17 -05:00
    fe28cce09e
  • Move frontmatter field table above examples in specification doc 
    Place the field constraints table immediately after the `Frontmatter`
heading so readers see the reference before the examples. Also remove
"(required)" from the heading, wrap examples in a `<Card>`, and relabel
"With optional fields:" to "Example with optional fields:".

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
    Jonathan Hefner · 2026-03-10 14:16:17 -05:00
    5d16d72435
  • Add SKILL.md filename annotations to code blocks in specification doc 
    The fenced code blocks showing `SKILL.md` content were labeled as `yaml`
or plain `markdown`. Changing them to `markdown SKILL.md` adds a
filename hint and more accurately reflects that they contain full
`SKILL.md` snippets (including `---` frontmatter delimiters), not just
raw YAML.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
    Jonathan Hefner · 2026-03-10 14:16:17 -05:00
    7a05447ecb
  • Wrap examples in <Card> components in specification doc 
    The `####` field headings didn't stand out enough amongst all the code
blocks, making the page hard to scan. Wrapping each field's examples in
a `<Card>` draws a border around them and adds padding, so each
subsection looks more contained instead of bleeding into the next. Also
bolds the example labels within each card.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
    Jonathan Hefner · 2026-03-10 14:16:16 -05:00
    a9d33af2d5
  • Expand directory tree in specification doc 
    Show optional directories (`scripts/`, `references/`, `assets/`)
directly in the directory tree with inline annotations, replacing the
separate `<Tip>` block.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
    Jonathan Hefner · 2026-03-10 14:16:16 -05:00
    cbc354c2ff
  • Improve clarity of name field constraints in specification doc 
    Place inline code examples next to the terms they describe (e.g.,
"`a-z`" next to "alphanumeric characters") rather than grouping them at
the end. Also add the word "hyphen" before the code-formatted `-` for
readability.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
    Jonathan Hefner · 2026-03-10 14:16:15 -05:00
    08d1d688c1
  • Format directory and file names as inline code in specification doc 
    Wrap `SKILL.md`, `scripts/`, `references/`, and `assets/` headings in
backticks for consistency with the rest of the document, which already
uses inline code formatting for these names in body text.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
    Jonathan Hefner · 2026-03-10 14:16:15 -05:00
    3736fdb995
  • Fix example description to include "when to use" guidance 
    The example `description` values in what-are-skills.mdx and
specification.mdx only described what the skill does, but omitted
when to use it. This contradicted the spec's own guidance that
description should cover both. Added a "Use when..." clause while
keeping the line short enough to avoid clipping in the rendered
code block on the docs site.
    Koichi ITO · 2026-03-10 10:41:26 +09:00
    22d6aeb353
  • Merge pull request #206 from jonathanhefner/add-optimizing-descriptions-guide 
    Add "Optimizing descriptions" guide for skill creators
    Jonathan Hefner · 2026-03-09 11:43:58 -05:00
    f2451a722f
  • Add instructionsUrl / sourceCodeUrl fields to LogoCarousel entries 
    Add documentation and source code URLs to each client's logo entry in
`LogoCarousel.jsx`. These fields aren't rendered yet, but serve as
structured metadata that future PRs can build on. Including them now
also encourages contributors adding new clients to provide these URLs,
and makes entries easier to verify during review.

Also reformats logo entries from single-line objects to multi-line for
readability.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
    Jonathan Hefner · 2026-03-05 14:42:40 -06:00
    caac642012
  • Add "Optimizing descriptions" guide for skill creators 
    How-to guide covering the full description optimization workflow:
writing effective descriptions, designing trigger eval queries
(should-trigger and should-not-trigger with near-miss examples), testing
trigger rates with a bash eval script, train/validation splits to avoid
overfitting, and the iterative optimization loop.

The guide is client-agnostic by default but includes a working Claude
Code example in the `check_triggered` function using
`--output-format json` and `jq` to detect `Skill` tool calls.

Adds the page to the "For skill creators" navigation group in
`docs.json`.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
    Jonathan Hefner · 2026-03-05 11:28:23 -06:00
    e9dff6b3d2
  • Merge pull request #203 from jonathanhefner/add-evaluating-skills-guide 
    Add "Evaluating skills" guide for skill creators
    Jonathan Hefner · 2026-03-05 11:24:53 -06:00
    09055204dd
  • Add comprehensive "Integrating skills" guide for client implementors (#200) 
    * Add comprehensive "Integrating skills" guide for client implementors

Replace the existing `integrate-skills.mdx` page with a thorough,
research-backed guide at `client-implementation/integrating-skills.mdx`.

The guide covers the full Agent Skills integration lifecycle in five
steps: discover, parse, disclose, activate, and manage context. It was
developed from analysis of seven real-world implementations (OpenCode,
Pi, Gemini CLI, Codex, VS Code Copilot Chat, Goose, OpenHands) and
addresses the practical design decisions developers face: progressive
disclosure architecture, cross-client `.agents/skills/`
interoperability, lenient YAML parsing for compatibility, catalog XML
format, behavioral instruction templates (derived from Pi's production
prompts), file-read vs. dedicated tool activation, structured wrapping
for context compaction, and permission allowlisting for bundled
resources.

The guide is architecture-neutral — it covers local filesystem agents,
cloud-hosted/sandboxed agents, and agents with or without file-read
capabilities, framing these as independent design choices rather than
rigid categories.

Also adds a redirect from the old `/integrate-skills` URL and updates
all internal links (`docs.json`, `home.mdx`, `what-are-skills.mdx`).

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

* Address PR review feedback on integrating skills guide

- Clarify that the XML catalog format is just one option — JSON or
  bulleted lists work equally well
- Qualify "works with any model" to note it requires a file-reading tool

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

* Rename "Integrating skills" guide to "Adding skills support"

Addresses reviewer feedback that "Integrating" could be misread as
integrating skills into a workflow rather than into a client. Renames
`integrating-skills.mdx` to `adding-skills-support.mdx` and updates all
references in `docs.json`, `home.mdx`, and `what-are-skills.mdx`.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
    Jonathan Hefner · 2026-03-05 11:24:08 -06:00
    07278fe14a
  • Add "Evaluating skills" guide for skill creators 
    A how-to guide for evaluating skill output quality using structured
evals. Covers the full eval workflow: designing test cases, running
with-skill vs. baseline comparisons, writing assertions, LLM-based
grading, aggregating benchmarks, analyzing patterns, human review, and
LLM-driven iterative improvement.

Derived from the workflow implemented by the `skill-creator` Skill, but
written as a standalone guide that readers can follow without using that
tool.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
    Jonathan Hefner · 2026-03-04 09:37:49 -06:00
    863f7c2857
  • Add Snowflake logo to Agent Skills logo carousel (#199) 
    .... Generated with [Cortex Code](https://docs.snowflake.com/en/user-guide/cortex-code/cortex-code)

Co-authored-by: Cortex Code <noreply@snowflake.com>
    Umesh Unnikrishnan · 2026-03-02 05:02:58 -08:00
    f019a02dbb
  • Add package.json to docs for local mint dev dependency (#141) 
    * Add package.json to docs for local mint dev dependency

Allow installing mint as a local dev dependency instead of requiring
a global install, making the setup more reproducible.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* Use npx mint dev to avoid mint version bug

Address PR review feedback: the mint package has a bug that causes it
to use the wrong version when installed as an npm dependency. Switch to
npx mint dev, remove the devDependency and package-lock.json, and
simplify docs accordingly.

* Move `package.json` to repo root and clean up docs

Move the `package.json` from `docs/` to the repo root so `npm run dev`
works from anywhere in the repository. Update `CLAUDE.md` instructions
to match, trim boilerplate from `docs/README.md`, and link to the live
documentation site.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
Co-authored-by: Jonathan Hefner <jonathan@hefner.pro>
    lilac · 2026-02-27 17:47:34 -06:00
    fbb6c829fd
  • Add "Using scripts" guide for skill creators (#196) 
    * Add "Using scripts" guide for skill creators

New guide at `docs/skill-creation/using-scripts.mdx` covering how to use
commands and scripts in skills:

- One-off commands with `uvx`, `pipx`, `npx`, `bunx`, `go run`,
  `deno run` (tabbed by ecosystem, with pinned version examples)
- Referencing bundled scripts from `SKILL.md` using relative paths
- Self-contained scripts with inline dependency declarations (PEP 723,
  Deno `npm:` imports, Bun auto-install, Ruby `bundler/inline` — tabbed
  with a common HTML-parsing example)
- Designing scripts for agentic use: non-interactive execution, `--help`
  documentation, error messages, structured output, and a compressed
  checklist of further considerations

Also updates `docs/docs.json` to organize navigation into groups ("For
skill creators" and "For client implementors").

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

* Address review feedback on "Using scripts" guide

Relative paths note: Clarify that the convention applies to support
files like `references/*.md`, and explain *why* (the agent runs commands
from the skill root).

Structured output: Reframe motivation around composability with both
agents and standard tools (`jq`, `cut`, `awk`) rather than LLM parsing
ambiguity. Shorten prose; let the code example's inline comments carry
the contrast.

Predictable output size: Add `--output` flag as an alternative strategy
for scripts whose output is large and not amenable to pagination. The
`--output` flag acts as a consent mechanism — the agent must explicitly
choose a file destination or pass `-` to opt in to stdout, preventing
accidental context-window flooding.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
    Jonathan Hefner · 2026-02-27 15:52:36 -06:00
    6102affec8