@warlock.js/ai-anthropic 5.34.1 → 5.34.3

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
package/CHANGELOG.md CHANGED
@@ -4,6 +4,18 @@ All notable changes to `@warlock.js/ai-anthropic` are documented in this file.
4
4
 
5
5
  The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.1.0/). `@warlock.js/*` packages are released in lockstep — every package shares the same version number, so a version below may list only the changes that affected this package.
6
6
 
7
+ ## 5.34.3 - 2026-10-20
8
+
9
+ ### Changed
10
+
11
+ - The agent skill now exports as `warlock-js-ai-anthropic`; package discovery descriptions stay within 450 characters.
12
+
13
+ ## 5.34.2 - 2026-10-19
14
+
15
+ ### Fixed
16
+
17
+ - Replace the loose skill catalog with a package index that points to the provider setup topic, and include orientation in generated LLM projections.
18
+
7
19
  ## 5.32.0 - 2026-10-18
8
20
 
9
21
  ### Changed
package/llms-full.txt CHANGED
@@ -4,11 +4,11 @@
4
4
 
5
5
  > Generated artifact. Concatenates every SKILL.md and reference file under `@warlock.js/ai-anthropic/skills/`. Re-run `node scripts/generate-llms.mjs` after any change.
6
6
 
7
- ## setup-anthropic `@warlock.js/ai-anthropic/setup-anthropic/SKILL.md`
7
+ ## ai-anthropic `@warlock.js/ai-anthropic/skills/SKILL.md`
8
8
 
9
9
  ---
10
- name: setup-anthropic
11
- description: 'Wire @warlock.js/ai-anthropic — new AnthropicSDK({apiKey, baseURL?, provider?}) for Claude, .model({name, vision?, structuredOutput?, reasoning?, thinkingMode?, promptCaching?, maxTokens?, maxOutputTokens?}). System-prompt hoisting, max_tokens required (default: model max output when streaming, 16384 non-streaming; maxOutputTokens for gateway ids), extended thinking via options.reasoning → adaptive thinking + output_config.effort (current/unknown models) or budget_tokens (Haiku 4.5, 4.6 and older; override with thinkingMode), prompt caching via promptCaching + options.cacheControl (per-call ttl), provider blocks (thinking/compaction) replayed verbatim via ModelResponse.blocks / Message.blocks, providerOptions.anthropic (betas, body fields such as context_management / fallbacks, thinkingDisplay), model.countTokens(), distinct finish reasons (pause_turn, refusal, context_exceeded), built-in Claude list prices (ANTHROPIC_MODEL_PRICING, anthropicPricingFor, withAnthropicCacheWriteRates), cost-truth usage (inputIncludesCache: false, cachedTokens/cacheWriteTokens/cacheWrite1hTokens, usage.cost on complete()), ProviderOverloadedError for 529, no first-party embeddings. Triggers: `AnthropicSDK`, `anthropic.model`, `anthropic.count`, `maxTokens`, `maxOutputTokens`, `max_tokens`, `reasoning`, `thinking`, `budget_tokens`, `thinkingMode`, `adaptive`, `cacheControl`, `providerOptions`, `blocks`, `countTokens`, `pause_turn`, `refusal`, `context_management`, `contextManagement`, `serverCompaction`, `clear_tool_uses`, `clear_thinking`, `thinkingBlockBinding`, `block_binding`, `preservedThinking`, `ai.providerTool`, `web_search_20250305`, `bash_20250124`, `text_editor_20250728`, `streamChunks`, `thinking_delta`, `compaction`, `promptCaching`, `cachedTokens`, `cacheWriteTokens`, `cacheWrite1hTokens`, `inputIncludesCache`, `ANTHROPIC_MODEL_PRICING`, `ProviderOverloadedError`, `claude-sonnet-4-6`, `claude-haiku-4-5`, `claude-opus-4-7`; "wire claude into warlock agent", "configure anthropic provider", "use claude sonnet", "anthropic gateway baseURL", "claude extended thinking", "anthropic prompt caching cost"; typical import `import { AnthropicSDK } from "@warlock.js/ai-anthropic"`. Skip: embeddings — the `setup-openai` topic of the `warlock-js-ai-openai` skill; sibling adapters `@warlock.js/ai-openai`, `@warlock.js/ai-bedrock`, `@warlock.js/ai-google`, `@warlock.js/ai-ollama`; raw `@anthropic-ai/sdk`; Vercel `@ai-sdk/anthropic`.'
10
+ name: warlock-js-ai-anthropic
11
+ description: 'Anthropic Claude adapter for Warlock AI: `AnthropicSDK`, `AnthropicModel`, `AnthropicCallOptions` and `ANTHROPIC_MODEL_PRICING`. Use for "connect Claude", streaming, tools, thinking, caching, native search or context management through `sdk.model()`. Not this package → @warlock.js/ai for agents and custom tools; @warlock.js/ai-bedrock for Claude on AWS; @warlock.js/ai-cli for terminal agents.'
12
12
  ---
13
13
 
14
14
  # `@warlock.js/ai-anthropic`
@@ -468,7 +468,7 @@ errors propagate. See the cost/budget guide for reserve mode.
468
468
 
469
469
  ## No embeddings
470
470
 
471
- Anthropic ships no first-party embeddings API, so `AnthropicSDK` intentionally does **not** implement `embedder()` (the `SDKAdapterContract.embedder` slot is optional). Use `@warlock.js/ai-openai` (or another embeddings-capable adapter; see the `setup-openai` topic of the `warlock-js-ai-openai` skill) for vectors.
471
+ Anthropic ships no first-party embeddings API, so `AnthropicSDK` intentionally does **not** implement `embedder()` (the `SDKAdapterContract.embedder` slot is optional). Use `@warlock.js/ai-openai` (or another embeddings-capable adapter; see the `warlock-js-ai-openai` skill) for vectors.
472
472
 
473
473
  ## When NOT to use this skill
474
474
 
package/llms.txt CHANGED
@@ -6,4 +6,4 @@
6
6
 
7
7
  ## Skills
8
8
 
9
- - [setup-anthropic](@warlock.js/ai-anthropic/setup-anthropic/SKILL.md): Wire @warlock.js/ai-anthropic — new AnthropicSDK({apiKey, baseURL?, provider?}) for Claude, .model({name, vision?, structuredOutput?, reasoning?, thinkingMode?, promptCaching?, maxTokens?, maxOutputTokens?}). System-prompt hoisting, max_tokens required (default: model max output when streaming, 16384 non-streaming; maxOutputTokens for gateway ids), extended thinking via options.reasoning → adaptive thinking + output_config.effort (current/unknown models) or budget_tokens (Haiku 4.5, 4.6 and older; override with thinkingMode), prompt caching via promptCaching + options.cacheControl (per-call ttl), provider blocks (thinking/compaction) replayed verbatim via ModelResponse.blocks / Message.blocks, providerOptions.anthropic (betas, body fields such as context_management / fallbacks, thinkingDisplay), model.countTokens(), distinct finish reasons (pause_turn, refusal, context_exceeded), built-in Claude list prices (ANTHROPIC_MODEL_PRICING, anthropicPricingFor, withAnthropicCacheWriteRates), cost-truth usage (inputIncludesCache: false, cachedTokens/cacheWriteTokens/cacheWrite1hTokens, usage.cost on complete()), ProviderOverloadedError for 529, no first-party embeddings. Triggers: `AnthropicSDK`, `anthropic.model`, `anthropic.count`, `maxTokens`, `maxOutputTokens`, `max_tokens`, `reasoning`, `thinking`, `budget_tokens`, `thinkingMode`, `adaptive`, `cacheControl`, `providerOptions`, `blocks`, `countTokens`, `pause_turn`, `refusal`, `context_management`, `contextManagement`, `serverCompaction`, `clear_tool_uses`, `clear_thinking`, `thinkingBlockBinding`, `block_binding`, `preservedThinking`, `ai.providerTool`, `web_search_20250305`, `bash_20250124`, `text_editor_20250728`, `streamChunks`, `thinking_delta`, `compaction`, `promptCaching`, `cachedTokens`, `cacheWriteTokens`, `cacheWrite1hTokens`, `inputIncludesCache`, `ANTHROPIC_MODEL_PRICING`, `ProviderOverloadedError`, `claude-sonnet-4-6`, `claude-haiku-4-5`, `claude-opus-4-7`; "wire claude into warlock agent", "configure anthropic provider", "use claude sonnet", "anthropic gateway baseURL", "claude extended thinking", "anthropic prompt caching cost"; typical import `import { AnthropicSDK } from "@warlock.js/ai-anthropic"`. Skip: embeddings — the `setup-openai` topic of the `warlock-js-ai-openai` skill; sibling adapters `@warlock.js/ai-openai`, `@warlock.js/ai-bedrock`, `@warlock.js/ai-google`, `@warlock.js/ai-ollama`; raw `@anthropic-ai/sdk`; Vercel `@ai-sdk/anthropic`.
9
+ - [ai-anthropic](@warlock.js/ai-anthropic/skills/SKILL.md): Anthropic Claude adapter for Warlock AI: `AnthropicSDK`, `AnthropicModel`, `AnthropicCallOptions` and `ANTHROPIC_MODEL_PRICING`. Use for "connect Claude", streaming, tools, thinking, caching, native search or context management through `sdk.model()`. Not this package → @warlock.js/ai for agents and custom tools; @warlock.js/ai-bedrock for Claude on AWS; @warlock.js/ai-cli for terminal agents.
package/package.json CHANGED
@@ -15,13 +15,13 @@
15
15
  "directory": "packages/ai-anthropic"
16
16
  },
17
17
  "peerDependencies": {
18
- "@warlock.js/ai": "5.34.1"
18
+ "@warlock.js/ai": "5.34.3"
19
19
  },
20
20
  "dependencies": {
21
21
  "@anthropic-ai/sdk": "^0.129.0",
22
- "@warlock.js/logger": "5.34.1"
22
+ "@warlock.js/logger": "5.34.3"
23
23
  },
24
- "version": "5.34.1",
24
+ "version": "5.34.3",
25
25
  "main": "./cjs/index.cjs",
26
26
  "module": "./esm/index.mjs",
27
27
  "types": "./esm/index.d.mts",
@@ -1,6 +1,6 @@
1
1
  ---
2
- name: setup-anthropic
3
- description: 'Wire @warlock.js/ai-anthropic — new AnthropicSDK({apiKey, baseURL?, provider?}) for Claude, .model({name, vision?, structuredOutput?, reasoning?, thinkingMode?, promptCaching?, maxTokens?, maxOutputTokens?}). System-prompt hoisting, max_tokens required (default: model max output when streaming, 16384 non-streaming; maxOutputTokens for gateway ids), extended thinking via options.reasoning → adaptive thinking + output_config.effort (current/unknown models) or budget_tokens (Haiku 4.5, 4.6 and older; override with thinkingMode), prompt caching via promptCaching + options.cacheControl (per-call ttl), provider blocks (thinking/compaction) replayed verbatim via ModelResponse.blocks / Message.blocks, providerOptions.anthropic (betas, body fields such as context_management / fallbacks, thinkingDisplay), model.countTokens(), distinct finish reasons (pause_turn, refusal, context_exceeded), built-in Claude list prices (ANTHROPIC_MODEL_PRICING, anthropicPricingFor, withAnthropicCacheWriteRates), cost-truth usage (inputIncludesCache: false, cachedTokens/cacheWriteTokens/cacheWrite1hTokens, usage.cost on complete()), ProviderOverloadedError for 529, no first-party embeddings. Triggers: `AnthropicSDK`, `anthropic.model`, `anthropic.count`, `maxTokens`, `maxOutputTokens`, `max_tokens`, `reasoning`, `thinking`, `budget_tokens`, `thinkingMode`, `adaptive`, `cacheControl`, `providerOptions`, `blocks`, `countTokens`, `pause_turn`, `refusal`, `context_management`, `contextManagement`, `serverCompaction`, `clear_tool_uses`, `clear_thinking`, `thinkingBlockBinding`, `block_binding`, `preservedThinking`, `ai.providerTool`, `web_search_20250305`, `bash_20250124`, `text_editor_20250728`, `streamChunks`, `thinking_delta`, `compaction`, `promptCaching`, `cachedTokens`, `cacheWriteTokens`, `cacheWrite1hTokens`, `inputIncludesCache`, `ANTHROPIC_MODEL_PRICING`, `ProviderOverloadedError`, `claude-sonnet-4-6`, `claude-haiku-4-5`, `claude-opus-4-7`; "wire claude into warlock agent", "configure anthropic provider", "use claude sonnet", "anthropic gateway baseURL", "claude extended thinking", "anthropic prompt caching cost"; typical import `import { AnthropicSDK } from "@warlock.js/ai-anthropic"`. Skip: embeddings — the `setup-openai` topic of the `warlock-js-ai-openai` skill; sibling adapters `@warlock.js/ai-openai`, `@warlock.js/ai-bedrock`, `@warlock.js/ai-google`, `@warlock.js/ai-ollama`; raw `@anthropic-ai/sdk`; Vercel `@ai-sdk/anthropic`.'
2
+ name: warlock-js-ai-anthropic
3
+ description: 'Anthropic Claude adapter for Warlock AI: `AnthropicSDK`, `AnthropicModel`, `AnthropicCallOptions` and `ANTHROPIC_MODEL_PRICING`. Use for "connect Claude", streaming, tools, thinking, caching, native search or context management through `sdk.model()`. Not this package → @warlock.js/ai for agents and custom tools; @warlock.js/ai-bedrock for Claude on AWS; @warlock.js/ai-cli for terminal agents.'
4
4
  ---
5
5
 
6
6
  # `@warlock.js/ai-anthropic`
@@ -460,7 +460,7 @@ errors propagate. See the cost/budget guide for reserve mode.
460
460
 
461
461
  ## No embeddings
462
462
 
463
- Anthropic ships no first-party embeddings API, so `AnthropicSDK` intentionally does **not** implement `embedder()` (the `SDKAdapterContract.embedder` slot is optional). Use `@warlock.js/ai-openai` (or another embeddings-capable adapter; see the `setup-openai` topic of the `warlock-js-ai-openai` skill) for vectors.
463
+ Anthropic ships no first-party embeddings API, so `AnthropicSDK` intentionally does **not** implement `embedder()` (the `SDKAdapterContract.embedder` slot is optional). Use `@warlock.js/ai-openai` (or another embeddings-capable adapter; see the `warlock-js-ai-openai` skill) for vectors.
464
464
 
465
465
  ## When NOT to use this skill
466
466
 
package/skills/README.md DELETED
@@ -1,9 +0,0 @@
1
- # `@warlock.js/ai-anthropic` — skills index
2
-
3
- Per-task skills. Cross-references name the topic ("the `<topic>` topic"), and for another package add its skill ("the `<topic>` topic of the `warlock-js-<pkg>` skill").
4
-
5
- ## Skills
6
-
7
- ### `setup-anthropic`
8
-
9
- Wire @warlock.js/ai-anthropic — new AnthropicSDK({apiKey, baseURL?, provider?}) for Claude, .model({name, vision?, structuredOutput?, reasoning?, thinkingMode?, promptCaching?, maxTokens?, maxOutputTokens?}). System-prompt hoisting, max_tokens required (default: model max output when streaming, 16384 non-streaming), extended thinking via options.reasoning → adaptive thinking + output_config.effort (current/unknown models) or budget_tokens (Haiku 4.5, 4.6 and older; override with thinkingMode), prompt caching via promptCaching + options.cacheControl with cost-truth usage (cachedTokens/cacheWriteTokens), no first-party embeddings. Load when wiring a Claude-backed model into a @warlock.js agent.