fix: add 'anthropic/' prefix to model pattern matching

[xlyoung]

Problem

Users with self-deployed Anthropic models use naming conventions like anthropic/claude-... (the standard LiteLLM routing format). The existing _matches_provider_pattern only recognized claude- and anthropic. prefixes, causing these models to be incorrectly filtered out when explicitly setting provider='anthropic'.

Fixes #5893

Fix

Added anthropic/ to the recognized prefix list for the anthropic provider in _matches_provider_pattern:

# Before
["claude-", "anthropic."]

# After
["claude-", "anthropic.", "anthropic/"]

This is backwards-compatible and handles the common LiteLLM model naming convention.

Summary by CodeRabbit

• Bug Fixes
  • Anthropic model provider detection now supports the anthropic/ model name prefix format, in addition to existing naming patterns.

[coderabbitai]

📝 Walkthrough

Walkthrough

Extended the Anthropic provider pattern matching in LLM._matches_provider_pattern to recognize models prefixed with anthropic/, in addition to the existing claude- and anthropic. prefixes. This enables CrewAI to correctly identify custom-deployed Anthropic models using the anthropic/ naming convention.

Changes

Anthropic Provider Pattern Matching

Layer / File(s)	Summary
Anthropic provider pattern matching extension lib/crewai/src/crewai/llm.py	Updated _matches_provider_pattern method to recognize the anthropic/ prefix when identifying models as belonging to the Anthropic provider, expanding support beyond the prior claude- and anthropic. patterns.

Estimated code review effort

🎯 1 (Trivial) | ⏱️ ~2 minutes

Poem

🐰 A model by any prefix to roam,
Whether claude-, anthropic., or anthropic/ home,
CrewAI now sees them all,
No more models left behind in the hall,
Provider patterns dance, wide and free! ✨

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title accurately summarizes the main change: adding the 'anthropic/' prefix to model pattern matching in the LLM provider matching logic.
Linked Issues check	✅ Passed	The PR directly addresses issue #5893 by adding 'anthropic/' to the recognized prefix list for Anthropic models, solving the reported filtering issue.
Out of Scope Changes check	✅ Passed	The changes are minimal and focused solely on updating the prefix pattern matching in LLM._matches_provider_pattern, directly aligned with the linked issue's requirements.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[coderabbitai]

🧹 Nitpick comments (2)

lib/crewai/src/crewai/llm.py (2)

114-114: ⚖️ Poor tradeoff

Consider using this constant to eliminate duplication.

The ANTHROPIC_PREFIXES constant is defined but not used in _matches_provider_pattern (line 459) or _is_anthropic_model (line 652), both of which define their own inline prefix lists. This leads to three different Anthropic prefix lists in the codebase:

• Line 114: ("anthropic/", "claude-", "claude/")
• Line 459: ["claude-", "anthropic.", "anthropic/"]
• Line 652: ("anthropic/", "claude-", "claude/")

These lists have different contents ("claude/" vs "anthropic."), which could cause subtle bugs where one method recognizes a model but another doesn't.

Consider consolidating to a single source of truth, or documenting why different prefix sets are needed for different purposes.

🤖 Prompt for AI Agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

In `@lib/crewai/src/crewai/llm.py` at line 114, Replace the three inline Anthropic
prefix lists with the single source-of-truth constant ANTHROPIC_PREFIXES: update
_matches_provider_pattern and _is_anthropic_model to reference
ANTHROPIC_PREFIXES (or a derived normalized set if a variant is required), and
reconcile the differing entries ("anthropic.", "claude/") by deciding the
canonical prefixes and documenting/normalizing them before use so all checks use
the same prefix definitions.

---

652-653: ⚡ Quick win

Consider using startswith instead of in to avoid false positives.

The substring match (prefix in model.lower()) on line 653 is less precise than the prefix match (model_lower.startswith(prefix)) used in _matches_provider_pattern (line 458). This could cause false positives:

• "my-provider/anthropic/claude-3" would incorrectly match as Anthropic
• "openai/custom-claude-like" would incorrectly match as Anthropic

Additionally, this method's prefix list ("anthropic/", "claude-", "claude/") differs from _matches_provider_pattern's list ["claude-", "anthropic.", "anthropic/"] — notably missing the "anthropic." prefix while including "claude/".

🔍 Proposed fix to use startswith and align with _matches_provider_pattern

 `@staticmethod`
 def _is_anthropic_model(model: str) -> bool:
     """Determine if the model is from Anthropic provider.
     
     Args:
         model: The model identifier string.
     
     Returns:
         bool: True if the model is from Anthropic, False otherwise.
     """
-    anthropic_prefixes = ("anthropic/", "claude-", "claude/")
-    return any(prefix in model.lower() for prefix in anthropic_prefixes)
+    model_lower = model.lower()
+    anthropic_prefixes = ("anthropic/", "claude-", "anthropic.")
+    return any(model_lower.startswith(prefix) for prefix in anthropic_prefixes)

🤖 Prompt for AI Agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

In `@lib/crewai/src/crewai/llm.py` around lines 652 - 653, The current check uses
substring matching which can yield false positives; in the function defining
anthropic_prefixes and returning any(prefix in model.lower() ...), change to use
a lower-cased model variable (e.g., model_lower = model.lower()) and use
model_lower.startswith(prefix) instead of `in`, and align the prefix list with
_matches_provider_pattern by using prefixes ["claude-", "anthropic.",
"anthropic/"] (remove "claude/" and add "anthropic."). Ensure the updated logic
still returns any(model_lower.startswith(prefix) for prefix in
anthropic_prefixes).

🤖 Prompt for all review comments with AI agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

Nitpick comments:
In `@lib/crewai/src/crewai/llm.py`:
- Line 114: Replace the three inline Anthropic prefix lists with the single
source-of-truth constant ANTHROPIC_PREFIXES: update _matches_provider_pattern
and _is_anthropic_model to reference ANTHROPIC_PREFIXES (or a derived normalized
set if a variant is required), and reconcile the differing entries
("anthropic.", "claude/") by deciding the canonical prefixes and
documenting/normalizing them before use so all checks use the same prefix
definitions.
- Around line 652-653: The current check uses substring matching which can yield
false positives; in the function defining anthropic_prefixes and returning
any(prefix in model.lower() ...), change to use a lower-cased model variable
(e.g., model_lower = model.lower()) and use model_lower.startswith(prefix)
instead of `in`, and align the prefix list with _matches_provider_pattern by
using prefixes ["claude-", "anthropic.", "anthropic/"] (remove "claude/" and add
"anthropic."). Ensure the updated logic still returns
any(model_lower.startswith(prefix) for prefix in anthropic_prefixes).

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: Organization UI

Review profile: CHILL

Plan: Pro Plus

Run ID: 66726fda-e2dd-4bc7-9084-7a2b702581e8

📥 Commits

Reviewing files that changed from the base of the PR and between 906cd97 and c8a43e4.

📒 Files selected for processing (1)

• lib/crewai/src/crewai/llm.py