<feed xmlns='http://www.w3.org/2005/Atom'>
<title>anthropic-skills.git, branch main</title>
<subtitle>Official agent skills collection
</subtitle>
<link rel='alternate' type='text/html' href='https://git.1991861.xyz/anthropic-skills.git/'/>
<entry>
<title>Update docx, pptx and xlsx skills: zipfile unpack, zip -y, validate.py and soffice.py fixes (#2221)</title>
<updated>2026-11-16T14:22:47+00:00</updated>
<author>
<name>Peter Lai</name>
<email>peterlai@anthropic.com</email>
</author>
<published>2026-11-16T14:22:47+00:00</published>
<link rel='alternate' type='text/html' href='https://git.1991861.xyz/anthropic-skills.git/commit/?id=e4bb5d960d87126ff3b3a8af1a104b7db67560ba'/>
<id>e4bb5d960d87126ff3b3a8af1a104b7db67560ba</id>
<content type='text'>
</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
</pre>
</div>
</content>
</entry>
<entry>
<title>Update docx, pptx, xlsx, and pdf skills (#2217)</title>
<updated>2026-11-14T15:16:42+00:00</updated>
<author>
<name>Peter Lai</name>
<email>peterlai@anthropic.com</email>
</author>
<published>2026-11-14T15:16:42+00:00</published>
<link rel='alternate' type='text/html' href='https://git.1991861.xyz/anthropic-skills.git/commit/?id=1780c5e5a552fd4d1c9d6a7b0cab26b0c2bef064'/>
<id>1780c5e5a552fd4d1c9d6a7b0cab26b0c2bef064</id>
<content type='text'>
</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
</pre>
</div>
</content>
</entry>
<entry>
<title>Move Office skills to the Apache 2.0 license (#2146)</title>
<updated>2026-11-06T22:32:28+00:00</updated>
<author>
<name>Mahesh Murag</name>
<email>mahesh@anthropic.com</email>
</author>
<published>2026-11-06T22:32:28+00:00</published>
<link rel='alternate' type='text/html' href='https://git.1991861.xyz/anthropic-skills.git/commit/?id=2c5958ecac7cee2b583967a2c4981d0f4a75c579'/>
<id>2c5958ecac7cee2b583967a2c4981d0f4a75c579</id>
<content type='text'>
docx, pdf, pptx and xlsx now use the same Apache 2.0 license as most other skills here, and the README is updated to match.</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
docx, pdf, pptx and xlsx now use the same Apache 2.0 license as most other skills here, and the README is updated to match.</pre>
</div>
</content>
</entry>
<entry>
<title>Update mcp-builder skill: teach v2 of the MCP TypeScript SDK (#2087)</title>
<updated>2026-10-28T19:20:38+00:00</updated>
<author>
<name>Felix Weinberger</name>
<email>3823880+felixweinberger@users.noreply.github.com</email>
</author>
<published>2026-10-28T19:20:38+00:00</published>
<link rel='alternate' type='text/html' href='https://git.1991861.xyz/anthropic-skills.git/commit/?id=ca45fc496a3436251f191354091d557da31c15b5'/>
<id>ca45fc496a3436251f191354091d557da31c15b5</id>
<content type='text'>
* Update mcp-builder skill: add a TypeScript SDK v2 guide

Adds reference/node_mcp_server_v2.md for @modelcontextprotocol/server, tells SKILL.md readers how to choose between v1 and v2, and links each version's docs. The v1 guide is unchanged apart from a note pointing new servers to v2.

* Update mcp-builder skill: tell v1 users that SDK v2 exists

When a TypeScript project already depends on @modelcontextprotocol/sdk, the agent tells the user once that v2 is the current stable release, that v1 gets only bug and security fixes, and how to migrate with the codemod, then continues with v1 unless asked to migrate.

* Update mcp-builder skill: rewrite the v2 guide around the SDK examples

Rebuilds reference/node_mcp_server_v2.md on patterns from the MCP TypeScript SDK examples: a server factory, a fetch-based API helper, tool design guidance, one complete stdio + HTTP example, bearer auth, resources and prompts, and a self-check script that verifies the built server. Links SDK example stories for advanced features. SKILL.md points v2 projects to the self-check.

* Update mcp-builder skill: review fixes for the TypeScript SDK v2 guide

---------

Co-authored-by: Konstantin Konstantinov &lt;konstantin@mach5technology.com&gt;</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
* Update mcp-builder skill: add a TypeScript SDK v2 guide

Adds reference/node_mcp_server_v2.md for @modelcontextprotocol/server, tells SKILL.md readers how to choose between v1 and v2, and links each version's docs. The v1 guide is unchanged apart from a note pointing new servers to v2.

* Update mcp-builder skill: tell v1 users that SDK v2 exists

When a TypeScript project already depends on @modelcontextprotocol/sdk, the agent tells the user once that v2 is the current stable release, that v1 gets only bug and security fixes, and how to migrate with the codemod, then continues with v1 unless asked to migrate.

* Update mcp-builder skill: rewrite the v2 guide around the SDK examples

Rebuilds reference/node_mcp_server_v2.md on patterns from the MCP TypeScript SDK examples: a server factory, a fetch-based API helper, tool design guidance, one complete stdio + HTTP example, bearer auth, resources and prompts, and a self-check script that verifies the built server. Links SDK example stories for advanced features. SKILL.md points v2 projects to the self-check.

* Update mcp-builder skill: review fixes for the TypeScript SDK v2 guide

---------

Co-authored-by: Konstantin Konstantinov &lt;konstantin@mach5technology.com&gt;</pre>
</div>
</content>
</entry>
<entry>
<title>Update claude-api skill: dynamic workflows and the four workflow quickstarts (#1995)</title>
<updated>2026-10-09T18:45:49+00:00</updated>
<author>
<name>CJ Avilla</name>
<email>cj@anthropic.com</email>
</author>
<published>2026-10-09T18:45:49+00:00</published>
<link rel='alternate' type='text/html' href='https://git.1991861.xyz/anthropic-skills.git/commit/?id=dbd4588f9e1033efb41dad4bef2f7947c8993d44'/>
<id>dbd4588f9e1033efb41dad4bef2f7947c8993d44</id>
<content type='text'>
Sync of the bundled claude-api skill: the Managed Agents dynamic workflows
guidance, the four dynamic-workflow quickstart templates, and the limited-
networking web-tool rules. Public-only trims in SKILL.md, shared/evals/ and
shared/model-migration.md are preserved.

Co-authored-by: Claude &lt;noreply@anthropic.com&gt;</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
Sync of the bundled claude-api skill: the Managed Agents dynamic workflows
guidance, the four dynamic-workflow quickstart templates, and the limited-
networking web-tool rules. Public-only trims in SKILL.md, shared/evals/ and
shared/model-migration.md are preserved.

Co-authored-by: Claude &lt;noreply@anthropic.com&gt;</pre>
</div>
</content>
</entry>
<entry>
<title>Update claude-api skill: Claude Haiku 5.5 as the current Haiku, Managed Agents and preserved-thinking fixes (#1991)</title>
<updated>2026-10-09T10:06:51+00:00</updated>
<author>
<name>Lance Martin</name>
<email>122662504+rlancemartin@users.noreply.github.com</email>
</author>
<published>2026-10-09T10:06:51+00:00</published>
<link rel='alternate' type='text/html' href='https://git.1991861.xyz/anthropic-skills.git/commit/?id=9d630808e4add0a7146de4af9384155d5dee350a'/>
<id>9d630808e4add0a7146de4af9384155d5dee350a</id>
<content type='text'>
</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
</pre>
</div>
</content>
</entry>
<entry>
<title>Update claude-api skill: managed-agents-onboard from a quickstart name or a URL (#1962)</title>
<updated>2026-10-05T13:46:42+00:00</updated>
<author>
<name>CJ Avilla</name>
<email>cj@anthropic.com</email>
</author>
<published>2026-10-05T13:46:42+00:00</published>
<link rel='alternate' type='text/html' href='https://git.1991861.xyz/anthropic-skills.git/commit/?id=683bc88e56f3e09ba94f7055977f3d3aa499f202'/>
<id>683bc88e56f3e09ba94f7055977f3d3aa499f202</id>
<content type='text'>
Co-authored-by: Claude Fable 5.1 &lt;noreply@anthropic.com&gt;</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
Co-authored-by: Claude Fable 5.1 &lt;noreply@anthropic.com&gt;</pre>
</div>
</content>
</entry>
<entry>
<title>Update claude-api skill: Claude Opus 5.5 default, Claude Sonnet 5.5, build-eval and hillclimb guides (#1930)</title>
<updated>2026-09-29T02:20:03+00:00</updated>
<author>
<name>Lance Martin</name>
<email>122662504+rlancemartin@users.noreply.github.com</email>
</author>
<published>2026-09-29T02:20:03+00:00</published>
<link rel='alternate' type='text/html' href='https://git.1991861.xyz/anthropic-skills.git/commit/?id=8a1541c4a3ffa5a20a5a91de0dcf3f0bab1d1ef4'/>
<id>8a1541c4a3ffa5a20a5a91de0dcf3f0bab1d1ef4</id>
<content type='text'>
* Update claude-api skill: Claude Opus 5.5 default, Claude Sonnet 5.5, build-eval and hillclimb guides

- Make Claude Opus 5.5 (`claude-opus-5-5`) the default model across SKILL.md,
  models.md, and every language's examples, with Claude Opus 5 as the previous
  Opus; add "Migrating to Claude Opus 5.5" to shared/model-migration.md
- Add Claude Sonnet 5.5 (`claude-sonnet-5-5`) as the current Sonnet, with
  Claude Sonnet 5 as the previous one, and update pricing, model-support
  callouts, and platform availability to match
- Add the build-eval and hillclimb flows under shared/evals/ (build-eval,
  eval-hillclimb, cost-hillclimb, eval-audit), the report schema, the lite HTML
  report builder, and the runner scaffold
- Add the preserved-thinking-migration subcommand with its cause reference and
  probe scripts, and note that drop_block can raise token usage
- Update cost-optimize for preserved thinking and per-model cache rates, and
  point cache misses at cache diagnostics
- Default streaming client tools to eager_input_streaming; document model
  access as a 404 and beta gating as a 400, not a 403
- Recommend `ant apply` for version-controlled Managed Agents resources; fix
  the thinking-binding default, Bedrock/Vertex header, and Managed Agents beta
  headers
- Use current-generation model ids in the Go, Java, and C# samples, and repoint
  dead live-sources URLs

* Update claude-api skill: drop references to files not shipped with the skill

- Remove the pointer to an examples directory that is not part of the skill
  from build-eval.md
- Remove the notes about a prebuilt report viewer's install source from
  SKILL.md, build-eval.md, and eval-hillclimb.md; the lite report builder
  remains the documented default
- Trim pre-launch behavior history from the thinking-block replay note and
  reword two "early-access" mentions in model-migration.md</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
* Update claude-api skill: Claude Opus 5.5 default, Claude Sonnet 5.5, build-eval and hillclimb guides

- Make Claude Opus 5.5 (`claude-opus-5-5`) the default model across SKILL.md,
  models.md, and every language's examples, with Claude Opus 5 as the previous
  Opus; add "Migrating to Claude Opus 5.5" to shared/model-migration.md
- Add Claude Sonnet 5.5 (`claude-sonnet-5-5`) as the current Sonnet, with
  Claude Sonnet 5 as the previous one, and update pricing, model-support
  callouts, and platform availability to match
- Add the build-eval and hillclimb flows under shared/evals/ (build-eval,
  eval-hillclimb, cost-hillclimb, eval-audit), the report schema, the lite HTML
  report builder, and the runner scaffold
- Add the preserved-thinking-migration subcommand with its cause reference and
  probe scripts, and note that drop_block can raise token usage
- Update cost-optimize for preserved thinking and per-model cache rates, and
  point cache misses at cache diagnostics
- Default streaming client tools to eager_input_streaming; document model
  access as a 404 and beta gating as a 400, not a 403
- Recommend `ant apply` for version-controlled Managed Agents resources; fix
  the thinking-binding default, Bedrock/Vertex header, and Managed Agents beta
  headers
- Use current-generation model ids in the Go, Java, and C# samples, and repoint
  dead live-sources URLs

* Update claude-api skill: drop references to files not shipped with the skill

- Remove the pointer to an examples directory that is not part of the skill
  from build-eval.md
- Remove the notes about a prebuilt report viewer's install source from
  SKILL.md, build-eval.md, and eval-hillclimb.md; the lite report builder
  remains the documented default
- Trim pre-launch behavior history from the thinking-block replay note and
  reword two "early-access" mentions in model-migration.md</pre>
</div>
</content>
</entry>
<entry>
<title>Update claude-api skill: link refusal billing to the docs (#1825)</title>
<updated>2026-09-24T16:20:35+00:00</updated>
<author>
<name>CJ Avilla</name>
<email>cj@anthropic.com</email>
</author>
<published>2026-09-24T16:20:35+00:00</published>
<link rel='alternate' type='text/html' href='https://git.1991861.xyz/anthropic-skills.git/commit/?id=33375500bcea98d610eb30ce10ac4e59b89c390d'/>
<id>33375500bcea98d610eb30ce10ac4e59b89c390d</id>
<content type='text'>
Point each refusal billing sentence at the "How refusals are billed"
docs section instead of restating the rules in the skill.

Also correct two claims to match that section: a refusal before any
output counts against rate limits, and a mid-stream refusal bills
input tokens as well as the output already streamed.</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
Point each refusal billing sentence at the "How refusals are billed"
docs section instead of restating the rules in the skill.

Also correct two claims to match that section: a refusal before any
output counts against rate limits, and a mid-stream refusal bills
input tokens as well as the output already streamed.</pre>
</div>
</content>
</entry>
<entry>
<title>Update claude-api skill: Managed Agents `auto` permission policy and `ant beta:sessions connect` (#1750)</title>
<updated>2026-09-10T19:44:08+00:00</updated>
<author>
<name>CJ Avilla</name>
<email>cj@anthropic.com</email>
</author>
<published>2026-09-10T19:44:08+00:00</published>
<link rel='alternate' type='text/html' href='https://git.1991861.xyz/anthropic-skills.git/commit/?id=34040c9c568585f6929bedeaad110ad08f079624'/>
<id>34040c9c568585f6929bedeaad110ad08f079624</id>
<content type='text'>
Adds the third Managed Agents permission policy, `auto`, alongside
`always_allow` / `always_ask`: the three outcomes (runs, denied as
high-risk with an error tool result while the session keeps running,
pauses for approval when indeterminate), a config example, what the
evaluation trusts, and the "not a human checkpoint" warning. Documents
the `evaluated_permission` and `evaluation` fields on `agent.tool_use` /
`agent.mcp_tool_use`, and updates the client-pattern and multiagent
guides to gate on `evaluated_permission === 'ask'` rather than the
configured policy.

Adds `ant beta:sessions connect` to the CLI guide: terminal viewer
keybindings, the allow/deny prompt, and the `--web` local session viewer.

Fixes the deny example to use `deny_message` (the real field) instead of
`message`.


Claude-Session: https://claude.ai/code/session_01UkZpc4FqLFPLBJF2Zcuq3a

Co-authored-by: Claude &lt;noreply@anthropic.com&gt;</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
Adds the third Managed Agents permission policy, `auto`, alongside
`always_allow` / `always_ask`: the three outcomes (runs, denied as
high-risk with an error tool result while the session keeps running,
pauses for approval when indeterminate), a config example, what the
evaluation trusts, and the "not a human checkpoint" warning. Documents
the `evaluated_permission` and `evaluation` fields on `agent.tool_use` /
`agent.mcp_tool_use`, and updates the client-pattern and multiagent
guides to gate on `evaluated_permission === 'ask'` rather than the
configured policy.

Adds `ant beta:sessions connect` to the CLI guide: terminal viewer
keybindings, the allow/deny prompt, and the `--web` local session viewer.

Fixes the deny example to use `deny_message` (the real field) instead of
`message`.


Claude-Session: https://claude.ai/code/session_01UkZpc4FqLFPLBJF2Zcuq3a

Co-authored-by: Claude &lt;noreply@anthropic.com&gt;</pre>
</div>
</content>
</entry>
</feed>
