rodin/review-bot
2
1
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases 6 Wiki Activity

feat: add Anthropic Messages API support (#18) #21

Merged
rodin merged 1 commits from feat/18-anthropic-api into main 2026-05-02 01:57:49 +00:00
Conversation 2 Commits 1 Files Changed 4
+247 -26
rodin commented 2026-05-02 01:49:26 +00:00
Owner
Copy Link
Copy Source

Adds --llm-provider flag (openai|anthropic) to switch between API formats.

Anthropic Implementation

  • POST /messages endpoint
  • x-api-key + anthropic-version headers
  • System prompt as top-level field (not a message)
  • max_tokens: 8192 for response generation
  • Parses content blocks [{type: "text", text: "..."}]

Changes

  • llm/client.go: Provider type, completeAnthropic(), doRequest() shared helper
  • cmd/review-bot/main.go: --llm-provider / LLM_PROVIDER flag
  • .gitea/actions/review/action.yml: llm-provider input + env
  • llm/client_test.go: 4 new tests for Anthropic path

Backwards compatible — default provider is still openai.

Closes #18

Adds `--llm-provider` flag (`openai`|`anthropic`) to switch between API formats. ## Anthropic Implementation - `POST /messages` endpoint - `x-api-key` + `anthropic-version` headers - System prompt as top-level field (not a message) - `max_tokens: 8192` for response generation - Parses content blocks `[{type: "text", text: "..."}]` ## Changes - `llm/client.go`: `Provider` type, `completeAnthropic()`, `doRequest()` shared helper - `cmd/review-bot/main.go`: `--llm-provider` / `LLM_PROVIDER` flag - `.gitea/actions/review/action.yml`: `llm-provider` input + env - `llm/client_test.go`: 4 new tests for Anthropic path Backwards compatible — default provider is still `openai`. Closes #18
rodin added 1 commit 2026-05-02 01:49:26 +00:00
feat: add Anthropic Messages API support (#18)
CI / test (pull_request) Successful in 13s
Details
CI / review (gpt-5, sonnet, SONNET_REVIEW_TOKEN) (pull_request) Successful in 1m2s
Details
CI / review (gpt-5-mini, gpt, GPT_REVIEW_TOKEN) (pull_request) Successful in 1m43s
Details
14a0c2a946 
Adds --llm-provider flag (openai|anthropic) to switch between API formats.

Anthropic implementation:
- POST /messages endpoint
- x-api-key + anthropic-version headers
- System prompt as top-level field (not a message)
- max_tokens: 8192 for response generation
- Parses content blocks [{type: "text", text: "..."}]

Changes:
- llm/client.go: Provider type, completeAnthropic(), doRequest() shared helper
- cmd/review-bot/main.go: --llm-provider / LLM_PROVIDER flag
- .gitea/actions/review/action.yml: llm-provider input + env
- llm/client_test.go: 4 new tests for Anthropic path

Backwards compatible — default provider is still openai.

Closes #18
sonnet-review-bot approved these changes 2026-05-02 01:50:42 +00:00
sonnet-review-bot left a comment
First-time contributor
Copy Link
Copy Source

Summary

Well-scoped addition of Anthropic Messages API support with clean provider abstraction and comprehensive tests. Defaults remain backwards-compatible and the action/CLI wiring is consistent.

Recommendation

APPROVE — The implementation cleanly introduces a provider switch with minimal disruption, adds appropriate headers and request/response handling for Anthropic, and covers the new paths with tests. The shared request helper improves code reuse and error reporting. CI has passed. Consider a future minor refinement to align the comment in completeAnthropic with its behavior (it currently uses the last system message encountered), but this is non-blocking. Proceed to merge.

## Summary Well-scoped addition of Anthropic Messages API support with clean provider abstraction and comprehensive tests. Defaults remain backwards-compatible and the action/CLI wiring is consistent. ## Recommendation **APPROVE** — The implementation cleanly introduces a provider switch with minimal disruption, adds appropriate headers and request/response handling for Anthropic, and covers the new paths with tests. The shared request helper improves code reuse and error reporting. CI has passed. Consider a future minor refinement to align the comment in completeAnthropic with its behavior (it currently uses the last system message encountered), but this is non-blocking. Proceed to merge.
gpt-review-bot approved these changes 2026-05-02 01:51:25 +00:00
gpt-review-bot left a comment
First-time contributor
Copy Link
Copy Source

Summary

The PR cleanly adds Anthropic Messages API support with a provider abstraction, shared HTTP handling, and tests. Changes are well-structured, tests cover new paths, and CI passed.

Findings

# Severity File Line Finding
1 [MINOR] llm/client.go 156 completeAnthropic hardcodes MaxTokens = 8192 in the request body. Consider making this configurable or document why this fixed limit is used (and whether it should be smaller depending on model).
2 [MINOR] llm/client.go 14 anthropic message mapping uses the same Message.Role/Message.Content fields as OpenAI. Verify this matches the exact Anthropic Messages API payload expectations (author vs role, content block shape). If the upstream API differs, the mapping may need adjustment.
3 [MINOR] cmd/review-bot/main.go 34 The new --llm-provider flag is validated and applied, but the flag parsing and usage order is fine. Consider documenting provider behavior in CLI help or README so users know provider-specific headers/requirements (e.g., anthropic requires anthropic-version and x-api-key).
4 [NIT] .gitea/actions/review/action.yml 34 The added default value line for llm-provider contains a trailing space after 'openai'. It's harmless but could be cleaned up for consistency.
5 [MINOR] llm/client_test.go 220 Some tests assert exact float equality when checking the presence/value of the temperature field decoded from JSON (e.g. comparing temp != 0.7). While likely fine, exact float comparisons can be brittle; consider checking presence and approximate equality or type before comparing.
6 [NIT] llm/client.go 220 doRequest now accepts a parse function and handles status checks after reading the body — this is good. Consider including the response status and a short snippet of the body in error messages only when safe (avoid leaking secrets) or add structured errors to aid higher-level handling.

Recommendation

APPROVE — Approve: the implementation is sound and the new provider abstraction is well integrated. Address the minor items to improve configurability and robustness: make Anthropic max_tokens configurable (or document the rationale), confirm the message payload shape exactly matches Anthropic's spec, optionally tighten tests that compare floats by value, and remove the trailing space in the action.yml default. Otherwise this is ready to merge.

## Summary The PR cleanly adds Anthropic Messages API support with a provider abstraction, shared HTTP handling, and tests. Changes are well-structured, tests cover new paths, and CI passed. ## Findings | # | Severity | File | Line | Finding | |---|----------|------|------|--------| | 1 | [MINOR] | `llm/client.go` | 156 | completeAnthropic hardcodes MaxTokens = 8192 in the request body. Consider making this configurable or document why this fixed limit is used (and whether it should be smaller depending on model). | | 2 | [MINOR] | `llm/client.go` | 14 | anthropic message mapping uses the same Message.Role/Message.Content fields as OpenAI. Verify this matches the exact Anthropic Messages API payload expectations (author vs role, content block shape). If the upstream API differs, the mapping may need adjustment. | | 3 | [MINOR] | `cmd/review-bot/main.go` | 34 | The new --llm-provider flag is validated and applied, but the flag parsing and usage order is fine. Consider documenting provider behavior in CLI help or README so users know provider-specific headers/requirements (e.g., anthropic requires anthropic-version and x-api-key). | | 4 | [NIT] | `.gitea/actions/review/action.yml` | 34 | The added default value line for llm-provider contains a trailing space after 'openai'. It's harmless but could be cleaned up for consistency. | | 5 | [MINOR] | `llm/client_test.go` | 220 | Some tests assert exact float equality when checking the presence/value of the temperature field decoded from JSON (e.g. comparing temp != 0.7). While likely fine, exact float comparisons can be brittle; consider checking presence and approximate equality or type before comparing. | | 6 | [NIT] | `llm/client.go` | 220 | doRequest now accepts a parse function and handles status checks after reading the body — this is good. Consider including the response status and a short snippet of the body in error messages only when safe (avoid leaking secrets) or add structured errors to aid higher-level handling. | ## Recommendation **APPROVE** — Approve: the implementation is sound and the new provider abstraction is well integrated. Address the minor items to improve configurability and robustness: make Anthropic max_tokens configurable (or document the rationale), confirm the message payload shape exactly matches Anthropic's spec, optionally tighten tests that compare floats by value, and remove the trailing space in the action.yml default. Otherwise this is ready to merge.
rodin merged commit 2adb23b3d9 into main 2026-05-02 01:57:49 +00:00
Sign in to join this conversation.
Reviewers
No Reviewers
sonnet-review-bot
gpt-review-bot
Labels
Clear labels
ai-review
bug
needs-review
ready
Ready for human review
 self-reviewed
Self-review passed since last push
 wip
Work in progress - do not start new work on this PR
No Label
Milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
security-review-bot rodin (Rodin)
No Assignees
3 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: rodin/review-bot#21
Delete Branch "feat/18-anthropic-api"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?