fix(flare-workflow): SDD-loop judge-parse retries + item_done in_review no-op

[getappz]

Summary

• build_sdd_loop_step's judge-parse-failure branch returned Ok(StepResult::Failure), but flare-workflow's execute_step_with_retry hardcodes should_retry=false for that variant — only a real Err consults the step's RetryPolicy. Every dispatched work-item run was failing outright on the first malformed judge reply, with zero retries, despite the step's RetryPolicy { max_attempts: 3, .. }. Confirmed against a live ~/.agentflare/workflows.db: 8/8 recent sdd_loop failures were single-attempt (attempt: 1, journaled as the generic "Step failed"). Now returns WorkflowError::StepFailed, which the engine does retry.
• item_done's in_review branch derived the response's status from a local bool instead of whether mark_in_review actually succeeded. mark_in_review silently no-ops (Ok(false)) on an ownership mismatch (claim lost/reclaimed), so a lost claim right after a successful push+PR was reported back as a clean "in_review" success — the item's state_id/assignee_agent never updated, leaving it looking orphaned and eligible for re-dispatch despite already having an open PR. Now surfaces this as a hard error.
• Added a throttled (5 min) claim heartbeat to run_or_resume_with_sender's poll loop — work-item jobs run up to WORK_JOB_TIMEOUT_SECS (~6h05m) but the claim TTL defaults to 30 minutes with nothing previously re-heartbeating it during a long SDD run.

Test plan

• cargo build --lib / full binary build
• cargo nextest run --bin agentflare work_item_pipeline (27/27 passed, including the updated judge_parse_failure_is_retryable_step_error)
• cargo nextest run --bin agentflare item_done (10/10 passed, incl. item_done_with_a_resulting_pr_moves_to_in_review_not_completed)
• cargo clippy --all-targets --all-features -- -D warnings -A unsafe_code -A clippy::pedantic
• cargo fmt --check

[coderabbitai]

Warning

Review limit reached

You’ve reached a temporary PR review limit under our Fair Usage Limits Policy.

Your recent review volume is higher than typical usage, so adaptive limits are currently applied.

Next review available in: 33 minutes

Enable usage-based reviews in Billing to review now. Otherwise, wait until the next included review is available.
You're only billed for reviews past your plan's rate limits ($0.25/file).

How can I continue?

After more reviews become available, a review can be triggered using the @coderabbitai review command as a PR comment. Alternatively, push new commits to this PR.

To avoid repeated limits, reduce automatic review volume by pausing incremental auto-reviews earlier, using label-based review opt-in, excluding WIP or generated PR titles, or requesting reviews manually when the PR is ready. If your team needs uninterrupted high-volume reviews, an organization admin can enable usage-based reviews.

How do review limits work?

CodeRabbit enforces per-developer PR review limits for each organization. Most developers receive the normal plan review availability.

For paid Pro and Pro+ PR reviews, CodeRabbit uses adaptive limits for sustained high-volume activity. When a developer's recent PR review activity reaches the 95th percentile or higher among CodeRabbit users, additional reviews become available more gradually as earlier reviews age out of the rolling window.

Please refer docs for additional details.

Review details

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Pro

Run ID: bcc5f972-dae5-4e88-9717-6fc3f420c0f9

📥 Commits

Reviewing files that changed from the base of the PR and between 7834aec and 7def449.

📒 Files selected for processing (2)

• src/mcp_server/item.rs
• src/work_item_pipeline.rs

---

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