feat: add /aidd-pr skill, rename /aidd-requirements, add eval infrastructure by ericelliott · Pull Request #168 · paralleldrive/aidd

ericelliott · 2026-03-24T22:58:39Z

Skill Split Notes

Per project standards (one skill per PR), the following skills from this PR have been split out into individual PRs. These split PRs should be merged first.

Skill	Split PR
`/aidd-parallel` (w/ delegate)	#187
`/aidd-pr`	#192
`/aidd-requirements`	#190
`/aidd-riteway-ai`	#189

Summary

Implements #166 and related housekeeping. Rebased on v3.0.0.

Changes

`/aidd-pr` skill

ai/skills/aidd-pr/SKILL.md — triage PR review comments, resolve addressed threads, delegate to /aidd-parallel
ai/skills/aidd-pr/README.md — man-page style reference
ai/commands/aidd-pr.md — command shortcut

`/aidd-parallel` skill

ai/skills/aidd-parallel/SKILL.md — generate /aidd-fix delegation prompts per task; /aidd-parallel delegate builds a mermaid dep graph and dispatches in dependency order
ai/skills/aidd-parallel/README.md — man-page style reference
ai/commands/aidd-parallel.md — command shortcut

`/aidd-requirements` rename

ai/skills/aidd-requirements/ — renamed from aidd-functional-requirements, includes man-page README
Updated references in aidd-task-creator, docs/learn-aidd-framework.md
ai/commands/aidd-requirements.md — new command file

`/aidd-riteway-ai` skill

ai/skills/aidd-riteway-ai/SKILL.md — 7-rule guide for writing correct .sudo prompt evals for multi-step tool-calling flows
ai/skills/aidd-riteway-ai/README.md — man-page style reference
ai/commands/aidd-riteway-ai.md — command shortcut
ai/skills/aidd-riteway-ai/riteway-ai.test.js — Vitest unit tests

Discovery

ai/skills/aidd-please/SKILL.md — all three new commands added to Commands block
AGENTS.md — review pull request => /aidd-pr added to Task Index
ai/skills/index.md — regenerated with new bullet format; includes all four new skills

Evals

ai-evals/aidd-pr/step-1-triage-test.sudo — tests observable triage reasoning (T_02 addressed, T_01 remaining)
ai-evals/aidd-pr/step-2-delegation-test.sudo — tests delegation prompt for T_01 only (consistent with step-1 output)
ai-evals/aidd-parallel/prompt-generation-test.sudo — tests /aidd-parallel prompt output
Fixture files colocated under each eval directory

CI

.github/workflows/ai-eval.yml — new dedicated workflow: daily at 8am UTC, path-filtered on ai-evals/**, continue-on-error: true
.github/workflows/test.yml — ai-eval job removed (now in ai-eval.yml)
package.json — test:ai-eval now globs ai-evals/**/*-test.sudo instead of a hardcoded path

Planned epics

tasks/aidd-parallel-skill-epic.md
tasks/aidd-riteway-ai-skill-epic.md
tasks/ai-eval-ci-epic.md

- Add ai/skills/aidd-pr/SKILL.md with 6-step PR triage process - Add ai/skills/aidd-pr/README.md with usage documentation - Add ai/commands/aidd-pr.md command shortcut - Update AGENTS.md Task Index with 'review pull request => /aidd-pr' - Regenerate index.md files to include new skill and command The /aidd-pr skill enables agents to: 1. Fetch PR details and review comments via gh CLI 2. Identify already-addressed comments and present for approval 3. Resolve approved conversations via GitHub GraphQL API 4. Validate remaining issues against current source 5. Generate focused /aidd-fix delegation prompts per issue 6. Optionally dispatch prompts to sub-agents Closes #166

ericelliott · 2026-03-25T01:13:40Z

@janhesters @ianwhitedeveloper any idea why the AI tests are failing?

Remove the 6-step SudoLang elaboration (Steps 1–6, the pipeline line, and the previous Constraints/Commands blocks). Insert the verbatim prompt from issue #166 directly after the Competencies block, preserving the frontmatter and preamble (role line + Competencies) unchanged.

… rename

…requirements

…ements, add unit/e2e distinction

…s for tool-calling skills

…kill

…delegation - Add ai/skills/aidd-parallel/SKILL.md with two commands: /aidd-parallel [--branch <branch>] <tasks> — generates one /aidd-fix delegation prompt per task, wrapped in a markdown codeblock with nested codeblocks indented, targeting the supplied branch directly /aidd-parallel delegate — builds file list + mermaid dep graph, sequences prompts by dependency order, and dispatches sub-agent workers - Add ai/skills/aidd-parallel/README.md with usage and workflow docs - Add ai/commands/aidd-parallel.md as an invokable command entry point - Add ai-evals/aidd-parallel/ with prompt-generation-test.sudo eval and fixtures - Update ai/skills/aidd-pr/SKILL.md to delegate prompt generation to /aidd-parallel and remove the inner Constraints block that now belongs in /aidd-parallel - Add /aidd-parallel to the Commands block in ai/skills/aidd-please/SKILL.md - Regenerate ai/ index files

…egration - Add ai/skills/aidd-riteway-ai/SKILL.md with 7 rules for writing correct riteway ai prompt evals for multi-step tool-calling flows - Add ai/commands/aidd-riteway-ai.md command file - Add unit tests covering frontmatter, rule content, command, and aidd-please integration (12 tests, all passing) - Add /aidd-riteway-ai to Commands block in ai/skills/aidd-please/SKILL.md - Regenerate ai/ index files - Include pending eval fixes for aidd-pr step-1 and greet fixture

ericelliott · 2026-03-30T10:36:03Z

@cursor please /review

cursor · 2026-03-30T10:36:08Z

That's the root cause. The test:ai-eval script only runs aidd-review/review-skill-test.sudo — but the failure is a Claude rate limit hit, not a test logic failure. The authentication token in CI has exhausted its quota. There's nothing broken in our code.

However, the test:ai-eval script also only covers the existing aidd-review eval — our new aidd-pr and aidd-parallel evals aren't wired into it yet. That's a separate gap worth noting but not causing this failure.

Summary for the PR: The CI ai-eval job is failing because "You've hit your limit · resets 7am UTC" — the Claude OAuth token in the repo's CI secrets has exhausted its usage quota. This is a credentials/quota issue, not a code problem. The test jobs (unit tests across Node 18/20/22) are all passing. Nothing in this PR broke the eval runner.

The question for @ericelliott/@janhesters/@ianwhitedeveloper: does the CI Claude account need its quota refreshed, or does the eval job need to be made non-blocking (continue-on-error: true) for rate-limit failures so they don't fail the PR?

…hedule

Add a test to lib/exports.test.js that verifies README.md references the renamed ai/skills/aidd-requirements/README.md path and does not contain the old broken ai/skills/aidd-functional-requirements/README.md link.

…append sections

cursor

Cursor Bugbot has reviewed your changes and found 2 potential issues.

^{❌ Bugbot Autofix is OFF. To automatically fix reported issues with cloud agents, enable autofix in the Cursor dashboard.}

^{Reviewed by Cursor Bugbot for commit c3b7eda. Configure here.}

cursor · 2026-04-06T19:12:43Z

+        uses: actions/upload-artifact@v4
+        with:
+          name: ai-eval-responses
+          path: ai-evals/*.responses.md


Artifact upload glob misses subdirectory response files

Medium Severity

The artifact upload path ai-evals/*.responses.md only matches files directly in the ai-evals/ directory, but all .sudo eval files live in subdirectories (ai-evals/aidd-pr/, ai-evals/aidd-parallel/, ai-evals/aidd-review/). Since --save-responses generates response files alongside the .sudo inputs, the glob never matches any responses. The path likely needs to be ai-evals/**/*.responses.md to capture files in subdirectories.

^{Reviewed by Cursor Bugbot for commit c3b7eda. Configure here.}

cursor · 2026-04-06T19:12:43Z

+
+fix bug => /aidd-fix
+review pull request => /aidd-pr`,
+    keywords: ["Task Index"],


Task Index append section uses wrong heading level

Medium Severity

The new Task Index entry in directiveAppendSections uses a ## heading, while every other entry in the array uses ###. When appendDirectives appends missing sections for upgrading users, they're nested under the ## AIDD Agent Directives (Auto-appended) wrapper. The ## heading makes Task Index appear as a sibling of the wrapper rather than a child, breaking the document hierarchy that all other append sections follow.

^{Reviewed by Cursor Bugbot for commit c3b7eda. Configure here.}

- Sub-agent isolation via /aidd-parallel delegate - Pipeline decomposed into effect and thinking subcommands - --output flag for arbitrary artifact types - Deterministic criteria auto-testing via Bun CLI - Blocked on PR #168 (aidd-parallel) Co-authored-by: Eric Elliott <support@paralleldrive.com>

Tracks #182 — creates tasks/aidd-critical-dependencies-epic.md listing the dependency chain (PR #168 → #181 → #184) and the order of operations for landing /aidd-parallel, /aidd-genesplice base, and genesplice epic. Co-authored-by: Eric Elliott <support@paralleldrive.com>

…endencies epic - Add PR #94 split plan: /aidd-rtc (3 files, zero deps), /aidd-upskill (11 files) - Add triage tasks for modified existing skills and infrastructure files - Update merge order: rtc → pipeline → parallel → pr → evals → upskill → genesplice - Document design decisions: parallel delegate is canonical, /aidd-delegate superseded - Split PR #168 into parallel, pr, and eval infrastructure tasks - Split PR #179: only /aidd-pipeline needed, /aidd-delegate excluded Co-authored-by: Eric Elliott <support@paralleldrive.com>

Concurrent dispatch + delegation skill extracted from PR #168. Includes the delegate subcommand, prompt generation, evals, and epic.

cursoragent added 20 commits March 25, 2026 01:40

feat(aidd-please): add /aidd-pr to Commands discovery block

dc55c7a

feat(aidd-pr): add delegation constraint to prevent direct execution

9120afb

fix(aidd-pr): move delegation constraint to orchestrator scope

1a0e5e7

fix(aidd-pr): soften delegation constraint to prefer over always

9d70a46

fix(aidd-pr): use original wording for delegation constraint

f4cebbe

docs(aidd-pr): rewrite README as man-page, tighten skill description

802105a

test(aidd-pr): add Riteway unit tests for skill structure and content

c9dc3bf

test(aidd-pr): add riteway ai prompt eval

0d801c1

test(aidd-pr): remove vitest structural tests replaced by ai eval

2b8b04a

test(aidd-pr): rewrite ai eval with fixture files and focused assertions

c6a248c

test(aidd-pr): supply mock gh and GraphQL tools in eval user prompt

fed13e9

test(aidd-pr): split eval into two focused step prompts

dd93f96

test(aidd-pr): step 1 should test tool calls, not pre-supply answers

f5918ff

plan(aidd-riteway-ai): add epic for riteway ai skill and requirements…

135589b

… rename

feat(aidd-requirements): rename aidd-functional-requirements to aidd-…

9244de7

…requirements

plan(aidd-riteway-ai): update epic - drop rename task, sharpen requir…

0d821c5

…ements, add unit/e2e distinction

plan(aidd-riteway-ai): note e2e infrastructure cost, prefer unit eval…

4c8e427

…s for tool-calling skills

revert(aidd-riteway-ai): undo premature e2e note, decision not yet made

1fdf942

plan(aidd-parallel): add epic for shared parallel prompt generation s…

b5ba856

…kill

cursor bot changed the title ~~feat(aidd-pr): add /aidd-pr skill for PR review and fix delegation~~ feat: add /aidd-pr skill, rename /aidd-requirements, add eval infrastructure Mar 30, 2026

cursoragent added 3 commits March 30, 2026 10:13

feat(commands): add aidd-requirements command file

a7891d3

ericelliott marked this pull request as ready for review March 30, 2026 10:35

Copilot AI review requested due to automatic review settings March 30, 2026 10:35

plan(ci): add epic for ai eval wiring, non-blocking job, and daily sc…

7ab2735

…hedule

Copilot AI review requested due to automatic review settings March 30, 2026 10:44

ericelliott review requested due to automatic review settings March 30, 2026 10:44

cursor bot reviewed Mar 30, 2026

View reviewed changes

Comment thread ai/skills/aidd-riteway-ai/SKILL.md

Comment thread ai-evals/aidd-pr/step-2-delegation-test.sudo Outdated

cursoragent added 2 commits March 30, 2026 11:01

fix(aidd-pr): clarify dangling branch reference in Constraints block

757ed7c

docs(aidd-riteway-ai): add man-page style README

4bd5039

Copilot AI review requested due to automatic review settings March 30, 2026 11:04

ericelliott review requested due to automatic review settings March 30, 2026 11:04

cursor bot reviewed Mar 30, 2026

View reviewed changes

Comment thread ai/skills/aidd-requirements/SKILL.md

Comment thread ai-evals/aidd-pr/step-1-triage-test.sudo Outdated

test: add failing test for aidd-requirements README link

9c7ba85

Add a test to lib/exports.test.js that verifies README.md references the renamed ai/skills/aidd-requirements/README.md path and does not contain the old broken ai/skills/aidd-functional-requirements/README.md link.

Copilot AI review requested due to automatic review settings March 31, 2026 21:10

ericelliott review requested due to automatic review settings March 31, 2026 21:10

cursor bot reviewed Mar 31, 2026

View reviewed changes

Comment thread package.json

feat(agents-md): add Task Index to requiredDirectives, template, and …

c3b7eda

…append sections

Copilot AI review requested due to automatic review settings April 6, 2026 19:07

ericelliott review requested due to automatic review settings April 6, 2026 19:07

cursor bot reviewed Apr 6, 2026

View reviewed changes

ericelliott mentioned this pull request Apr 8, 2026

docs: create plan/pending-prs.md with prioritized PR analysis #183

Open

ericelliott mentioned this pull request Apr 10, 2026

Add genesplice task epic for remaining work #184

Draft

ericelliott mentioned this pull request Apr 10, 2026

docs: add critical dependencies epic (PR #168 → #181 → #184) #185

Draft

ericelliott mentioned this pull request Apr 10, 2026

docs: add PR #94 split plan and updated landing order to critical dependencies epic #186

Draft

cursor bot pushed a commit that referenced this pull request Apr 10, 2026

feat(skills): add /aidd-parallel

df30bfb

Concurrent dispatch + delegation skill extracted from PR #168. Includes the delegate subcommand, prompt generation, evals, and epic.

This was referenced Apr 10, 2026

feat(skills): add /aidd-parallel #187

Merged

feat(skills): add /aidd-riteway-ai #189

Merged

feat(skills): add /aidd-requirements #190

Merged

feat(skills): add /aidd-pr #192

Merged

janhesters pushed a commit that referenced this pull request Apr 15, 2026

feat(skills): add /aidd-parallel

cdbff5f

Concurrent dispatch + delegation skill extracted from PR #168. Includes the delegate subcommand, prompt generation, evals, and epic.

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

feat: add /aidd-pr skill, rename /aidd-requirements, add eval infrastructure#168

feat: add /aidd-pr skill, rename /aidd-requirements, add eval infrastructure#168
ericelliott wants to merge 30 commits intomainfrom
cursor/aidd-config-json-support-24c1

ericelliott commented Mar 24, 2026 •

edited by cursor bot

Loading

Uh oh!

ericelliott commented Mar 25, 2026

Uh oh!

ericelliott commented Mar 30, 2026

Uh oh!

cursor bot commented Mar 30, 2026 •

edited

Loading

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

cursor bot left a comment

Uh oh!

cursor bot Apr 6, 2026

Uh oh!

cursor bot Apr 6, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

Conversation

ericelliott commented Mar 24, 2026 • edited by cursor bot Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Skill Split Notes

Summary

Changes

/aidd-pr skill

/aidd-parallel skill

/aidd-requirements rename

/aidd-riteway-ai skill

Discovery

Evals

CI

Planned epics

Uh oh!

ericelliott commented Mar 25, 2026

Uh oh!

ericelliott commented Mar 30, 2026

Uh oh!

cursor bot commented Mar 30, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

cursor bot left a comment

Choose a reason for hiding this comment

Uh oh!

cursor bot Apr 6, 2026

Choose a reason for hiding this comment

Artifact upload glob misses subdirectory response files

Uh oh!

cursor bot Apr 6, 2026

Choose a reason for hiding this comment

Task Index append section uses wrong heading level

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

ericelliott commented Mar 24, 2026 •

edited by cursor bot

Loading

`/aidd-pr` skill

`/aidd-parallel` skill

`/aidd-requirements` rename

`/aidd-riteway-ai` skill

cursor bot commented Mar 30, 2026 •

edited

Loading