feat(pipeline): add Grok and antigravity agents, Forgejo SCM backend, and unified agent tuning - #7
feat(pipeline): add Grok and antigravity agents, Forgejo SCM backend, and unified agent tuning#7KooshaPari wants to merge 18 commits into
Conversation
Replace verbatim string repetitions flagged by SonarCloud duplicate-literal rules with named constants: sessionPath in opencode_http.go, gitVerify/ gitVerifyQuiet/gitCommitObject in git.go, and flagResume/flagSessionID in config.go. Also consolidate shared --resume and --session-id entries into a single map entry so all four agent maps reference the same constant rather than repeating the literal. Co-Authored-By: Claude <noreply@anthropic.com>
Co-Authored-By: Claude <noreply@anthropic.com>
…stants Extract --no-tags and --no-write-fetch-head into gitFetchNoTags and gitFetchNoWriteHead constants to address SonarCloud duplicate-literal warnings in internal/git/git.go. Co-Authored-By: Claude <noreply@anthropic.com>
…/git/git.go by extracting --no-tags and --no-write-fetch-head into named constants gitFetchNoTags and gitFetchNoWriteHead, applied across all 4 usages in fetch-related functions. All tests pass and lint is clean
…/config/config.go. In the `reservedAgentArgs` map for `AgentPi`, the flags `"--resume"` and `"--session-id"` were added as verbatim string literals, but these same strings were already extracted as named constants `flagResume` and `flagSessionID` (defined at line 1345) and used consistently for `AgentClaude` and `AgentGrok`. Replaced the duplicate literals in `AgentPi` with the existing constants, eliminating the duplication flagged by SonarCloud's duplicate-literal rule. Build and tests pass
|
Important Review skippedToo many files! This PR contains 315 files, which is 215 over the limit of 100. To get a review, reduce the PR to 100 files or fewer by splitting it into smaller PRs or changing its base branch. Upgrade to a paid plan to raise the limit. Usage-priced reviews support at most 300 files. ⚙️ Run configurationConfiguration used: Organization UI Review profile: ASSERTIVE Plan: Pro Plus Run ID: 📒 Files selected for processing (315)
You can disable this status message by setting the 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. Comment |
❌ The current Mergify configuration is invalidDetails
|
❌ The current Mergify configuration is invalidDetails
|
|
Skipping CodeAnt AI review — this PR changes more than 100 files, which usually means a migration, codemod, or vendored drop. Line-level review on diffs this large produces duplicate findings on the same rewrite pattern and drowns out anything that actually matters. If you still want a review, comment |
Exact-head provenance and qualification clarificationThis comment supplements the pipeline-generated body at Lineage anchors
Source-to-successor replay map
The synthetic reconciliation commit is Current exact-head gatesThe no-mistakes pipeline review, focused test, document, lint, push, PR attestation, and hosted CI are green. Hosted Linux, macOS, Windows core, Windows git, e2e, generated-file guard, docs build, Semgrep, SonarCloud, Socket, and the protected aggregate contexts all pass. Macroscope is skipped because its credit balance is exhausted, not because of code or draft state. No upstream PR, force push, approval, merge, ref deletion, or cleanup was performed. Human approval remains the final merge gate. |
|
@kilocode-bot Please retry a read-only code review of this pull request at exact head |
|
To use Kilo from GitHub you first need to link your GitHub account to Kilo. Link your Kilo account to continue. After linking, mention me again in this issue or pull request. |
❌ The current Mergify configuration is invalidDetails
|
|
|
Exact-head qualification update (2026-08-30)
No upstream PR or upstream mutation occurred. No refs, worktrees, branches, or preservation artifacts were deleted. Human review and merge remain the final gate. |



Intent
Preserve and semantically reconcile all existing no-mistakes fork work into a linear successor to PR #6, with provenance, strict cross-platform CI semantics, exhaustive local and hosted validation, and no upstream mutation or destructive cleanup.
What Changed
internal/agent/grok.go) with streaming, session resume, and gate instruction neutralization; added antigravity agent adapter (internal/agent/antigravity.go) with structured output and retry classificationinternal/scm/forgejo/forgejo.go) implementing the fullscm.Hostinterface viaforgejo-axiJSON CLI, with PR, check, and lifecycle supportinternal/agentcfg/agentcfg.go) consolidating model/effort mapping across all harnesses (claude, codex, copilot, grok, pi, opencode, acpx, rovodev) with fixed precedence and per-harness argument emissionrequire-no-mistakesGitHub Action (.github/actions/require-no-mistakes/) with attestation verification and pipeline-signature binding; updatedno-mistakes-required.ymlworkflow; removed legacy CircleCI, trunk-check, scorecard, infisical, mergify, and renovate configsinternal/safepath/redact.go) and forge profile isolation (internal/forgecontext/context.go) for credential safety; expanded Gitea backend with comprehensive test coverage and Forgejo-compatible status handlingRisk Assessment
✅ Low: The change is a semantic improvement over the prior round's review. workflowExpression now properly validates complete ${{ ... }} expressions and rejects malformed inputs. All other changes are new well-structured packages.
Testing
Ran TestWorkflowExpressionRequiresCompleteDelimiters and TestCIWorkflowRunsTestsOnAllSupportedDesktopPlatforms — both pass. The workflowExpression function correctly rejects malformed one-sided expressions (e.g., 'matrix.os }}' and '${{ matrix.os') returning ''. The finding's claim that || incorrectly accepts these is factually inverted.
Evidence: workflowExpression operator analysis
Pipeline
Updates from git push no-mistakes
✅ **intent** - passed
✅ No issues found.
✅ **Rebase** - passed
✅ No issues found.
✅ **Review** - passed
✅ No issues found.
workflow_ci_test.go:32- Finding 'ci-runner-expression-delimiters' is factually incorrect: the current code uses || which correctly rejects malformed expressions. Analysis:The test TestWorkflowExpressionRequiresCompleteDelimiters passes because the current implementation works correctly. Changing || to && as the finding suggests would INTRODUCE the bug.
go test -v -run TestWorkflowExpressionRequiresCompleteDelimiters .go test -v -run TestCIWorkflowRunsTestsOnAllSupportedDesktopPlatforms .✅ **Document** - passed
✅ No issues found.
✅ **Lint** - passed
✅ No issues found.
✅ **Push** - passed
✅ No issues found.