Add Crater ForgeFed OpenAI compatibility provider

[konard]

Summary

• add UPSTREAM_PROVIDER=crater with CRATER_FORGEFED_INBOX, actor, target, and poll configuration
• translate OpenAI chat completion requests into ForgeFed Offer activities containing Ticket objects, parse Accept.result, poll task URIs until resolution, and map results back to OpenAI-compatible JSON or SSE
• introduce the async TaskProvider trait so the Crater backend can be swapped without changing the HTTP route layer
• document Crater setup and add a changelog fragment

Reproduce / Verify

1. Start the router with a Crater inbox, for example:

   UPSTREAM_PROVIDER=crater \
   CRATER_FORGEFED_INBOX=https://tracker.example/inbox \
   link-assistant-router serve

2. POST OpenAI-style chat JSON to /v1/chat/completions.
3. Use either body "stream": true or /v1/chat/completions?stream=true to receive SSE chunks.

Tests

• cargo fmt --all -- --check
• cargo clippy --all-targets --all-features --verbose
• rust-script scripts/check-file-size.rs
• cargo test --all-features --verbose
• cargo test --doc --verbose
• rust-script scripts/check-changelog-fragment.rs

Fixes #33

[konard]

Working session summary

Implemented and pushed the Crater ForgeFed provider in PR 34:

#34

Summary:

• Added UPSTREAM_PROVIDER=crater with Crater inbox/actor/target/poll config.
• Added OpenAI-compatible /v1/chat/completions handling for sync and SSE, including ?stream=true.
• Translates chat requests into ForgeFed Offer + Ticket, parses Accept.result, polls task resolution, and maps back to OpenAI-style responses.
• Added async TaskProvider abstraction, tests, docs, and changelog fragment.

Verification:

• Local checks passed: fmt, clippy, file-size check, all tests, doc tests, changelog check.
• PR CI passed on commit 8465363: lint/format, changelog, Ubuntu/macOS/Windows tests, and package build.
• PR is ready for review and the working tree is clean.

---

This summary was automatically extracted from the AI working session output.

[konard]

🤖 Solution Draft Log

This log file contains the complete execution trace of the AI solution draft process.

💰 Cost estimation:

• Model: GPT-5.5
• Provider: OpenAI
• Public pricing estimate: $21.516757

📊 Context and tokens usage:

• 761.0K / 1.1M (72%) input tokens, 46.7K / 128K (36%) output tokens

Total: (761.0K + 11.8M cached) input tokens, 46.7K output tokens, $21.516757 cost

🤖 Models used:

• Tool: OpenAI Codex
• Requested: gpt-5.5
• Model: GPT-5.5 (gpt-5.5)

📎 Log file uploaded as Repository (75629KB)

• View complete solution draft log

---

Now working session is ended, feel free to review and add any feedback on the solution draft.

[konard]

🎉 Auto-merged

This pull request has been automatically merged by hive-mind.

• All CI checks have passed

---

Auto-merged by hive-mind with --auto-merge flag