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 <noreply@anthropic.com>
* Update claude-api skill: Managed Agents self-hosted memory stores and web tool domain settings, cost-optimize subcommand, Admin API reference, ASCII-only text
Managed Agents: self-hosted sandboxes can now attach memory stores via the SDK worker (handle_item, ANTHROPIC_WORK_SECRET, sync options, troubleshooting); web_search/web_fetch accept allowed_domains/blocked_domains/user_location/max_content_tokens on the toolset configs entry and are not governed by environment networking; typed per-tool config unions; Console session viewer notes; packages caveat under limited networking; Claude Platform on AWS self-hosted worker auth.
New shared/cost-optimization.md backing a cost-optimize subcommand, and new shared/admin-api.md covering client.beta.organization in all SDKs and the CLI. Prompt caching gains TTL selection, automatic vs explicit breakpoint guidance, workspace isolation, and verification guidance. Sonnet 5 pricing is the permanent $2/$10 list price. Advisor pairing no longer excludes Claude Fable 5 for Managed Agents. Reviewer-only HTML comments are stripped from the published files.
All files are now plain ASCII in prose (em dashes, arrows, emoji callouts, and box-drawing replaced with ASCII equivalents), matching the source so future syncs diff cleanly.
No-Verification-Needed: documentation-only change to skill reference content
* Update claude-api skill: Claude Fable 5.1 / Mythos 5.1 catalog rows, Files and Skills APIs out of beta
Claude Fable 5.1 (claude-fable-5-1) and Claude Mythos 5.1 (claude-mythos-5-1) become the default Fable-tier models throughout the skill; Claude Fable 5 and Mythos 5 stay selectable by id with their own catalog rows. Feature-support lists that named Fable 5 now read Fable 5/5.1.
The Files API and Skills API are out of beta: examples use client.files.* / client.skills.* with no beta header, and the API-drift table points at the beta-to-GA migration docs.
No-Verification-Needed: documentation-only change to skill reference content
* Update claude-api skill: Claude Fable 5.1 / Mythos 5.1 migration section and API changes
Adds a "Migrating to Claude Fable 5.1 from Claude Fable 5" section to shared/model-migration.md: three breaking changes (forced tool_choice any/tool returns 400; thinking blocks are preserved only for the model that produced them or a newer one; and only in the conversation that produced them, so edited history replayed with thinking blocks is rejected), what carries over from Fable 5, the Opus 5 path, Mythos 5.1 notes, capability improvements, prompt-tunable behavioral shifts, and a migration checklist.
New API features documented: per-message effort (mid-conversation-output-config beta), turn-scoped mid-conversation system messages with clear_at, progress updates between tool calls via thinking.display "updates", thinking block_binding controls, and the 0.025x cache-read rate on Fable 5.1 with a max_tokens: 0 keep-alive that usually beats the 1-hour TTL.
Error catalog, prompt-caching, tool-use, platform-availability, cost-optimization, and prompt-audit are updated to match; SKILL.md routes migration and prompting questions to the new section.
No-Verification-Needed: documentation-only change to skill reference content
* Update claude-api skill: Claude Fable 5.1 launch-day hedges and migration-path table rows
Adds claude-fable-5 -> claude-fable-5-1 and claude-mythos-5 -> claude-mythos-5-1 rows to the migration-path and model-ID mapping tables (including the Bedrock IDs), and updates the refusal-fallback example to the 5.1 model id.
Hedges three claims until the launch docs confirm them: Task Budgets support on Claude Fable 5.1, whether Claude Mythos 5.1 shares the 0.025x cache-read rate, and the fallback-credit wording. The block_binding error row now says to send the controls beta header only where that beta is offered and to fall back to strip-and-retry elsewhere. Cross-references within the migration section point at the history-editing check directly.
No-Verification-Needed: documentation-only change to skill reference content
Adds an `upgrade` subcommand that walks a project from anthropic 0.x to
1.x (httpx2 HTTP layer, Python 3.10 floor, removed deprecated surface).
- New python/claude-api/sdk-upgrade.md: scope and version check, a
greppable inventory of affected call sites, one section per breaking
change, verification, and a report format.
- SKILL.md: `upgrade` subcommand row and a reading-guide entry.
- python/claude-api/README.md: timeouts use anthropic.Timeout, custom
clients note httpx2, link to the SDK MIGRATION.md.
- shared/live-sources.md: SDK major-version upgrade guides table.
Two changes to make the skill packageable as an uploaded custom skill,
with no other content changes:
- Rename: the skill folder, the SKILL.md frontmatter name, and the
marketplace.json entry name and path. Skill names cannot contain the
reserved words "claude" or "anthropic" (per the agent skills
best-practices documentation).
- Description: shortened from 1,176 to 992 characters to fit the
1,024-character limit that skill upload validation applies to the
SKILL.md frontmatter description. The shortened text is the version
tested internally.
Opt-in skill that appends two or three short, targeted follow-up questions after a substantive answer the user may act on, helping them check key facts, probe the reasoning and assumptions, and notice missing context before relying on the answer. Fires at most once per conversation, with explicit skip rules for trivial lookups, formatting tasks, code, and creative writing.
Opt-in skill that recommends matching Claude Academy (academy.claude.com) courses, tutorials, and use cases when a user asks how to use Claude or a Claude product. Reads the live catalog at academy.claude.com/assets/data/catalog.json at runtime rather than embedding a snapshot, so recommendations never go stale; when the catalog cannot be fetched it names no specific item and points to the relevant Academy product hub or the resources library instead. Only recommends on a strong match and never invents Academy content.
* Add prompt-audit subcommand to claude-api skill
New shared/prompt-audit.md: a non-interactive audit that inventories the
prompt surface, dates each instruction, scans against dated-pattern
tables, and produces a findings report plus a proposed diff. SKILL.md
gains the subcommand row, a cross-reference from migrate, carve-outs in
the provider check and language detection for the non-interactive
contract, and a reading-guide entry. model-migration.md points at the
audit from its prompt-behavior section.
No-Verification-Needed: markdown-only skill content, no runtime surface
* Replace prompt-audit exception callouts with an intent check in language detection
No-Verification-Needed: markdown-only skill content, no runtime surface
Mirror the latest bundled-skill content for the August 5 Managed Agents
launch wave:
- Session budgets: budget object at session create (minor-unit cents
string), budget_reached pause semantics, settle-event allowlist,
session.usage event, change/remove-to-resume, multiagent shared cap
- Deployment budgets: same budget object on deployments, copied onto
each fired session, clearable and re-addable
- Inference geo pinning: inference_geo on the agent model object,
per-session override, roster uniformity, never grandfathered
- Skills from a GitHub repository: root .claude/skills discovery at
session start (cloud sandboxes only), trust-boundary warning
- Advisor: {type: "advisor", model} multiagent roster entry, reserved
anthropic.advisor thread, plaintext vs redacted delivery; advisor
tool max_uses/max_tokens/caching options and error result variant
- Multiagent: new when-to-use guidance (start with self, add cheaper
workers, dedicated specialists), delegation tools, docs URL rename
- Corrections: agent versions are sequential integers, Files API
uploads take a purpose param, stop_details refusal categories are an
open set, Sonnet 5 added to the prefill-removed list, vault_ids is
create-only on sessions, deployments gain an update endpoint
No-Verification-Needed: markdown-only skill content sync
* Update claude-api skill: Claude Opus 5
- Add Claude Opus 5 (`claude-opus-5`) as the default model across all
SDK examples, model tables, pricing, and migration guidance.
- Add an Opus 5 migration section covering the API changes that come
with it, and move Opus 4.8 into the previous-generation slot.
- Drop the Opus 4.7 fast-mode guidance, which no longer applies.
- Scope the server-side fallbacks beta header to the current version.
* Update claude-api skill: Opus 5 fast-follow corrections
- Server-side refusal fallbacks are available on Claude Platform on AWS,
not just the Claude API. Correct the availability statements in
SKILL.md, the migration guide, and the platform-availability table.
- Simplify the effort guidance: start at `high` (the API default) and
sweep down, rather than starting at `xhigh` for coding work.
- Add a time-to-first-token section with the prompt instruction that
reduces pre-answer thinking on latency-sensitive routes.
- Replace the two separate thinking-disabled mitigations with the single
combined instruction that covers both failure modes.
Syncs the claude-api skill with the current upstream source.
## Managed Agents — five new features
- `effort` on the agent's `model` object (a level string or `{"type": "<level>"}`).
It is agent-configuration only: setting it in a per-session `model` override is
silently ignored.
- Optional `version` on agent update, for optimistic concurrency. Omit it for
last-write-wins.
- `initial_events` on session create, collapsing create plus first send into one
call. Validation is all-or-nothing and only `user.message` and
`user.define_outcome` are accepted.
- Environment and memory-store webhooks: four `environment.*` events and three
`memory_store.*` events.
- Event deltas on per-thread streams. Previews are thread-scoped, so a child
thread's previews never reach the session-level stream.
## Corrections to behavior the skill already documented
- Tool output offload triggers at 100,000 characters (~25k tokens), not 100K
tokens, and covers built-in tools rather than MCP alone.
- `system.message` works on four models, checks only the primary model, appends
system context instead of replacing the prompt, and is accepted during a
`requires_action` idle when it trails a tool result in the same request.
- Vault-to-MCP credential matching is normalized (scheme and host lowercased,
default ports and trailing slashes stripped), not byte-exact.
- Multiagent depth greater than 1 is a validation error, not silently ignored.
- Outcome `interrupted` fires even when evaluation never started, and then
carries an empty-string `outcome_evaluation_start_id`.
- Deployment jitter is 15% of the run interval, floor 5s, cap 9 minutes.
- Webhooks retry three times with jittered 5-120s backoff, then drop silently.
Auto-disable is duration-based with three named triggers.
- `session.status_terminated` means completion or error.
- `agent.thinking` is a progress signal and carries no thinking content.
- `processed_at` is already populated on first sighting for
`user.define_outcome`, `user.custom_tool_result`, and `user.tool_result`.
- Session creation does not provision the sandbox.
- Skill `version` applies to Anthropic-authored skills too.
- Console-created vault credentials are header-injection only, and vault
environment-variable substitution skips secrets in URL paths.
- Console trace URLs need the real workspace ID when the API key is not in the
Default workspace.
## Elsewhere in the skill
- Note partner pricing under the first-party price table: Microsoft Foundry
bills at standard API rates, while Amazon Bedrock and Vertex AI are
partner-operated with separate pricing.
- Correct the tool-runner human-in-the-loop guidance, which previously pointed
readers at the manual loop for approval gates the runner's per-turn hooks
already cover.
- Repoint links away from the retired documentation hosts.
- Dedupe eagerly loaded SKILL.md content that the per-section files already
cover.
Add support for template formats (.dotx, .potx, .xltx) across validation
and the helper scripts.
Consolidate the shared office helpers into a single module, replacing the
pack/unpack pipeline with explicit zip/unzip steps. Extraction now rejects
symlink and path-traversal archive entries. Move run merging to a
standalone docx script, since it only ever applied to Word documents.
Fix the redlining validator so it compares against the original even when
a document has no tracked changes, which is when an untracked edit would
otherwise go unreported.
Provision a LibreOffice user profile per invocation so conversions work in
sandboxed environments.
Trim the skill docs to the guidance that earns its place.
- Add Managed Agents scheduled deployments: new
shared/managed-agents-scheduled-deployments.md (cron schedules, deployment
runs, pause/auto-pause), Deployments and Deployment Runs API reference,
beta-header coverage, and SKILL.md routing
- Vault environment_variable credentials: secrets substituted at egress with
networking allowlists; secrets guidance in tools, client-patterns, and
onboarding docs rewritten around them; self-hosted sandbox caveats noted
- Add system.message event for mid-session system prompt updates (Opus 4.8
only) to the events guide and API reference
- Add Claude Fable 5 and Claude Mythos 5 to the model tables with pricing,
context window, and model-selection guidance
- Document Fable-specific API behavior: always-on adaptive thinking (explicit
disabled returns 400), protected-thinking display and replay rules, new
tokenizer (~30% more tokens), refusal stop reason with server-side fallbacks
and SDK fallback middleware, 30-day data-retention requirement
- Add the full Migrating to Claude Fable 5 guide section and checklist,
including the Mythos Preview migration path
- Refresh skill trigger description, effort/compaction/task-budget notes,
caching minimums, structured-output and dynamic-filtering support tables
The model field on agent create accepts {id, speed}, not
{type: "model_config", id, speed}. Aligns with the shape documented
elsewhere in managed-agents-core.md and the request body table.
Add shared/managed-agents-memory.md covering the Memory Stores public
beta under managed-agents-2026-04-01: object model (memstore_/mem_/
memver_), create + seed, attach via resources[] at session-create time,
FUSE mount at /mnt/memory/<store>/, host-side CRUD with create-by-path
vs update-by-id, content_sha256 preconditions, and versions/redact.
Wire it through all cross-references: SKILL.md beta-headers namespace
and reading guide; api-reference SDK method rows, delete/archive quirks
bullet, AddResource note, and three new endpoint sections; core.md
architecture diagram and resources[] enumerations; environments.md
Resources intro; overview.md beta-headers table, Reading Guide row, and
archive-is-permanent bullet.
* Add proper front-matter to SKILL.md for claude-api
Added metadata section with name, description, and license information.
* Apply suggestion from @ericharmeling
---------
Co-authored-by: Eric Harmeling <eric.harmeling@outlook.com>
* Update claude-api skill with Managed Agents guidance
* Replace OPUS_ID placeholder with concrete model string in claude-api skill
* Replace remaining model placeholders with concrete model names and IDs
improve_description.py now calls `claude -p`
as a subprocess instead of the Anthropic SDK, so users no longer need a
separate ANTHROPIC_API_KEY to run the description optimization loop. Same
auth pattern run_eval.py already used for the triggering eval.
Prompts go over stdin (they embed the full SKILL.md body). Strips CLAUDECODE
env var to allow nesting inside a Claude Code session. The over-1024-char
retry is now a fresh single-turn call that inlines the too-long version
rather than a multi-turn followup.
SKILL.md: dropped the stale "extended thinking" reference to match.
Documentation skill for building applications with the Claude API
and Agent SDK. Covers Python, TypeScript, Java, Go, Ruby, C#, PHP,
and cURL with language-specific guides for:
- Messages API basics, streaming, and error handling
- Tool use (tool runner and manual agentic loop)
- Structured outputs and adaptive thinking
- Batches and Files APIs
- Agent SDK patterns (Python/TypeScript)
- Model catalog and selection guidance
- Add `compatibility` optional field to SKILL.md frontmatter spec
- Add validation for `compatibility` field in quick_validate.py
- Rename "hyphen-case" to "kebab-case" terminology in init_skill.py
and quick_validate.py
- Update max skill name length from 40 to 64 characters
- Make scripts executable (chmod +x) for accept_changes.py,
comment.py, extract_form_structure.py, add_slide.py, thumbnail.py,
and recalc.py
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
docx: Add commenting and track-changes support. Reorganize OOXML
tooling into a shared office/ module.
pptx: Streamline SKILL.md, add slide-editing and pptxgenjs guides,
bundle html2pptx as a tgz. Reorganize OOXML tooling into a shared
office/ module.
xlsx: Move recalc script into scripts/ and expand it. Add shared
office/ module for OOXML pack/unpack/validate.
pdf: Improve form-filling workflow with new form-structure extraction
script and updated field-info extraction.
Added a note at the top of the README directing users to
agentskills.io for information about the Agent Skills standard.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-authored-by: Claude <noreply@anthropic.com>
Reorganize the spec documentation:
- agent-skills-spec.md now serves as an index linking to the guides
- skill-authoring.md covers skill creation and SKILL.md format
- skill-client-integration.md provides guidance for Skill Client implementors
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-authored-by: Claude <noreply@anthropic.com>
