feat(mcp): expose outcome calibration via gittensory_get_outcome_calibration by jason020818 · Pull Request #1174 · JSONbored/loopover

jason020818 · 2026-06-24T05:20:16Z

Summary

Add gittensory_get_outcome_calibration MCP tool (replaces closed feat(mcp): expose outcome calibration via gittensory_get_outcome_calibration #1171)
Guard slop summary with optional chaining so sparse reports cannot throw

Test plan

vitest mcp-output-schemas outcome_calibration test

Supersedes #1171 — addresses AI reviewer null-safety feedback.

…bration Wire the slop-band and recommendation outcome calibration service into MCP so maintainers can inspect predictive accuracy without a separate HTTP call. Includes structured output schema validation tests. Co-authored-by: Cursor <cursoragent@cursor.com>

Use optional chaining on the slop calibration slice so the MCP tool returns a safe fallback summary instead of throwing on sparse reports. Co-authored-by: Cursor <cursoragent@cursor.com>

superagent-security · 2026-06-24T05:20:43Z

Superagent didn't find any vulnerabilities or security issues in this PR.

codecov · 2026-06-24T05:22:25Z

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 94.86%. Comparing base (f085f78) to head (7cdec26).
⚠️ Report is 13 commits behind head on main.

Additional details and impacted files

@@           Coverage Diff           @@
##             main    #1174   +/-   ##
=======================================
  Coverage   94.85%   94.86%           
=======================================
  Files         156      156           
  Lines       18909    18920   +11     
  Branches     6848     6850    +2     
=======================================
+ Hits        17937    17948   +11     
  Misses        387      387           
  Partials      585      585

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

ghost · 2026-06-24T06:09:35Z

Tip

🟩🟩🟩🟩🟩🟩🟩🟩🟩🟩🟩🟩

✅ Gittensory review — safe to merge

3 files · 2 AI reviewers · no blockers · readiness 100/100 · CI green

✅ Approved — safe to merge

Review summary
The change adds a new MCP tool `gittensory_get_outcome_calibration` that returns slop‑band calibration data for a repository. It introduces input and output Zod schemas, registers the tool, and implements a private handler that builds the report and crafts a human‑readable summary. The implementation compiles and integrates with existing tooling, and tests are added to verify schema compliance and basic functionality.

Nits (5)

src/mcp/server.ts: The ternary `input.windowDays !== undefined ? input.windowDays : undefined` is redundant; passing `input.windowDays` directly would be clearer.
src/mcp/server.ts: The summary construction uses nested ternary operators, which reduces readability; consider extracting into a helper function.
src/mcp/server.ts: The output schema types `slop` and `recommendations` are `z.unknown()`. More specific schemas would improve validation and documentation.
test/unit/mcp-output-schemas.test.ts: The test expects `data.windowDays` to be present in the returned report; ensure `buildRepoOutcomeCalibration` includes this field or augment the handler to add it.
test/unit/mcp-output-schemas.test.ts: The test checks that `data.signals` is an array; verify that the calibration report always provides a `signals` array to satisfy the schema.

Signal	Result	Evidence
Code review	✅ No blockers	2 reviewers, synthesized
Linked issue	✅ Linked	#1171
Related work	✅ No active overlap found	No same-issue or scoped active PR overlap found.
Review load	✅ 20/20	Readiness component derived from cached public PR metadata and labels; size label size:M.
Validation evidence	✅ 25/25	PR body includes validation/test evidence.
Open PR queue	✅ 10/10	0 open PR(s), 0 likely reviewable.
Contributor context	✅ Confirmed Gittensor contributor	jaso0n0818; Gittensor profile; 1361 PR(s), 5 issue(s).
Gate result	✅ Passing	No configured blocker found.

Review context

Author: jaso0n0818
Role context: outside_contributor
Public audience mode: oss maintainer
Lane context: Repository is configured for direct PR review.
Public profile languages: Python, Rust, Shell, C++, Dart, Ruby, Scala, TypeScript
Official Gittensor activity: 1361 PR(s), 5 issue(s).
PR-specific overlap: none found.

Contributor next steps

Keep the PR focused and include validation evidence before maintainer review.

Signal definitions

Related work = same linked issue, overlapping active PRs, or title/path similarity.
Review load = cached public PR metadata such as size labels, changed paths, and preflight status.
Open PR queue = repo-wide review pressure; it is not a PR quality failure.
Contributor context = public GitHub/Gittensor identity context; non-Gittensor status is not a blocker.

Review details

Generated from public PR metadata and the diff. Advisory only; deterministic signals remain authoritative.

The change adds a new MCP tool `gittensory_get_outcome_calibration` that returns slop‑band calibration data for a repository. It introduces input and output Zod schemas, registers the tool, and implements a private handler that builds the report and crafts a human‑readable summary. The implementation compiles and integrates with existing tooling, and tests are added to verify schema compliance and basic functionality.

Nits (5)

src/mcp/server.ts: The ternary `input.windowDays !== undefined ? input.windowDays : undefined` is redundant; passing `input.windowDays` directly would be clearer.
src/mcp/server.ts: The summary construction uses nested ternary operators, which reduces readability; consider extracting into a helper function.
src/mcp/server.ts: The output schema types `slop` and `recommendations` are `z.unknown()`. More specific schemas would improve validation and documentation.
test/unit/mcp-output-schemas.test.ts: The test expects `data.windowDays` to be present in the returned report; ensure `buildRepoOutcomeCalibration` includes this field or augment the handler to add it.
test/unit/mcp-output-schemas.test.ts: The test checks that `data.signals` is an array; verify that the calibration report always provides a `signals` array to satisfy the schema.

_{🟩 Safe / merged · 🟦 Advisory · 🟨 Held for review · 🟥 Blocked / closed}

💰 Earn for open-source contributions like this. Gittensor lets GitHub contributors earn for the work they already do — register to start earning →.

Checked by Gittensory, a quiet PR intelligence layer for OSS maintainers.

Re-run Gittensory review

…ure helper

…intainer_noise (#1414) The deterministic maintainer-noise report (`buildMaintainerNoiseReport`, src/signals/reward-risk.ts) ranks which queue-noise sources a maintainer should clear first — open PRs without linked-issue context, broad/churn-style diffs, duplicate clusters, and contributor-intake pressure — with a score, level, and recommended maintainer actions. It already powers the `@gittensory noise-report` PR command but had no MCP surface, so agents/CLI couldn't pull it. Add `gittensory_get_maintainer_noise` (maintainer-authenticated, advisory only), mirroring the outcome-calibration serving (#1174): a thin load-or-compute service (`src/services/maintainer-noise.ts`) that loads the repo's cached signals and runs the existing builder, plus the MCP tool registration + output schema + method. No new REST route, no migration, no auth-policy change — purely additive to the MCP surface. Output flows through the existing `redactSensitiveForMcp` wrapper. Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com> Co-authored-by: ghost <49853598+JSONbored@users.noreply.github.com>

…r_lane The deterministic maintainer-lane report (`buildMaintainerLaneReport`, src/signals/engine.ts) synthesizes a repo's lane recommendation alongside the configured maintainer cut, queue health, config quality, and contributor-intake health — the maintainer's own-lane triage view. It already powers the repo-intelligence response but had no MCP surface, so agents/CLI couldn't pull it (it is also absent from gittensory_get_repo_context). Add `gittensory_get_maintainer_lane` (maintainer-authenticated via the existing requireRepoAccess, advisory only), mirroring the merged maintainer-noise / outcome-calibration serving (JSONbored#1414 / JSONbored#1174): a thin load-or-compute service (src/services/maintainer-lane.ts) that loads the repo's cached signals + collision report and runs the existing builder, plus the MCP tool registration + output schema + method. Output flows through the existing redactSensitiveForMcp wrapper. No new REST route, no migration, no auth-policy change — purely additive to the MCP surface. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

The deterministic label audit (`buildLabelAudit`, src/signals/engine.ts) reports whether a repo's configured (.gittensory.yml / dashboard) label set matches the live GitHub labels and is trustworthy for label-multiplier scoring — surfacing missing configured labels, suspicious status/source-style labels, and the overall trusted-label-pipeline readiness. It already powers the repo-intelligence response but had no MCP surface, so agents/CLI couldn't pull it (it is also absent from gittensory_get_repo_context, which exposes a different subset). Add `gittensory_get_label_audit` (maintainer-authenticated via the existing requireRepoAccess, advisory only), mirroring the merged maintainer-noise (JSONbored#1414) and outcome-calibration (JSONbored#1174) serving: a thin load-or-compute service (src/services/label-audit.ts) that loads the repo's labels + cached signals and runs the existing builder, plus the MCP tool registration + output schema + method. Output flows through the existing redactSensitiveForMcp wrapper. No new REST route, no migration, no auth-policy change — purely additive to the MCP surface. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

…r_lane (#1456) The deterministic maintainer-lane report (`buildMaintainerLaneReport`, src/signals/engine.ts) synthesizes a repo's lane recommendation alongside the configured maintainer cut, queue health, config quality, and contributor-intake health — the maintainer's own-lane triage view. It already powers the repo-intelligence response but had no MCP surface, so agents/CLI couldn't pull it (it is also absent from gittensory_get_repo_context). Add `gittensory_get_maintainer_lane` (maintainer-authenticated via the existing requireRepoAccess, advisory only), mirroring the merged maintainer-noise / outcome-calibration serving (#1414 / #1174): a thin load-or-compute service (src/services/maintainer-lane.ts) that loads the repo's cached signals + collision report and runs the existing builder, plus the MCP tool registration + output schema + method. Output flows through the existing redactSensitiveForMcp wrapper. No new REST route, no migration, no auth-policy change — purely additive to the MCP surface. Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com> Co-authored-by: ghost <49853598+JSONbored@users.noreply.github.com>

…dit (#1461) The deterministic label audit (`buildLabelAudit`, src/signals/engine.ts) reports whether a repo's configured (.gittensory.yml / dashboard) label set matches the live GitHub labels and is trustworthy for label-multiplier scoring — surfacing missing configured labels, suspicious status/source-style labels, and the overall trusted-label-pipeline readiness. It already powers the repo-intelligence response but had no MCP surface, so agents/CLI couldn't pull it (it is also absent from gittensory_get_repo_context, which exposes a different subset). Add `gittensory_get_label_audit` (maintainer-authenticated via the existing requireRepoAccess, advisory only), mirroring the merged maintainer-noise (#1414) and outcome-calibration (#1174) serving: a thin load-or-compute service (src/services/label-audit.ts) that loads the repo's labels + cached signals and runs the existing builder, plus the MCP tool registration + output schema + method. Output flows through the existing redactSensitiveForMcp wrapper. No new REST route, no migration, no auth-policy change — purely additive to the MCP surface. Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com> Co-authored-by: ghost <49853598+JSONbored@users.noreply.github.com>

jason020818 and others added 2 commits June 24, 2026 05:09

fix(mcp): guard outcome-calibration summary when slop data is absent

9be17a8

Use optional chaining on the slop calibration slice so the MCP tool returns a safe fallback summary instead of throwing on sparse reports. Co-authored-by: Cursor <cursoragent@cursor.com>

jason020818 requested a review from JSONbored as a code owner June 24, 2026 05:20

dosubot Bot added the size:M label Jun 24, 2026

ghost added gittensory:reviewed gittensor:feature Gittensor-scored feature linked to a feature issue — scores a 0.25x multiplier. labels Jun 24, 2026

refactor(mcp): derive the outcome-calibration summary from a tested p…

7cdec26

…ure helper

JSONbored merged commit d728e32 into JSONbored:main Jun 24, 2026
14 checks passed

github-actions Bot mentioned this pull request Jun 24, 2026

MCP release due: 0.7.0 #629

Closed

12 tasks

oktofeesh1 mentioned this pull request Jun 26, 2026

feat(mcp): expose maintainer queue-noise triage via gittensory_get_maintainer_noise #1414

Merged

23 tasks

oktofeesh1 mentioned this pull request Jun 27, 2026

feat(mcp): expose the repo label-policy audit via gittensory_get_label_audit #1602

Closed

This was referenced Jul 2, 2026

MCP release due: 1.0.0 joaovictor91123/gittensory#2

Open

MCP release due: 0.7.0 RealDiligent/gittensory#6

Open

Navigation Menu

Uh oh!

feat(mcp): expose outcome calibration via gittensory_get_outcome_calibration - #1174

feat(mcp): expose outcome calibration via gittensory_get_outcome_calibration#1174
JSONbored merged 3 commits into
JSONbored:mainfrom
jason020818:feat/mcp-outcome-calibration-v2

jason020818 commented Jun 24, 2026

Uh oh!

superagent-security Bot commented Jun 24, 2026

Uh oh!

codecov Bot commented Jun 24, 2026 •

edited

Loading

Uh oh!

ghost commented Jun 24, 2026 •

edited by ghost

Loading

✅ Gittensory review — safe to merge

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

Uh oh!

Conversation

jason020818 commented Jun 24, 2026

Summary

Test plan

Uh oh!

superagent-security Bot commented Jun 24, 2026

Uh oh!

codecov Bot commented Jun 24, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Codecov Report

Uh oh!

ghost commented Jun 24, 2026 • edited by ghost Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

✅ Gittensory review — safe to merge

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

codecov Bot commented Jun 24, 2026 •

edited

Loading

ghost commented Jun 24, 2026 •

edited by ghost

Loading