name: behavior-conformance-instructions
description: >
  Advisory-tier behavior conformance evals for 44 high-signal Markdown-applyTo
  instructions in `.github/instructions/**` (covers ADO, coding-standards,
  experimental, GitHub, hve-core, Jira, RAI planning, security, and shared
  scopes). Each stimulus prompts the model
  to identify the instruction that applies in a given working context and
  asserts both `applyTo` path evidence and instruction-specific scope
  vocabulary. Per DD-05, this is an advisory tier and grader false-positives
  are absorbed by design.
type: capability
defaults:
  runs: 3
  timeout: 120s
  executor: copilot-sdk

stimuli:
  - name: instruction-ado-backlog-sprint-conformance
    prompt: |
      You are preparing sprint planning artifacts under
      `.copilot-tracking/workitems/sprint/iter-42/`. Which
      `.github/instructions/**/*.instructions.md` file applies (cite its path)
      and summarize its top three requirements?
    tags:
      category: behavior-conformance
      instruction: ado-backlog-sprint
      advisory: "true"
    graders:
      - type: output-matches
        name: applyTo-evidence
        config:
          pattern: "(?i)\\.copilot-tracking/workitems/sprint|workitems/sprint/"
      - type: output-matches
        name: scope-language
        config:
          pattern: "(?i)sprint|iteration|capacity|coverage\\s+analysis"

  - name: instruction-ado-backlog-triage-conformance
    prompt: |
      You are about to triage a batch of untriaged Azure DevOps work items
      under `.copilot-tracking/workitems/triage/q4-cleanup/`. Which
      `.github/instructions/**/*.instructions.md` file applies (cite its path)
      and summarize the triage requirements it enforces?
    tags:
      category: behavior-conformance
      instruction: ado-backlog-triage
      advisory: "true"
    graders:
      - type: output-matches
        name: applyTo-evidence
        config:
          pattern: "(?i)\\.copilot-tracking/workitems/triage|workitems/triage/"
      - type: output-matches
        name: scope-language
        config:
          pattern: "(?i)triage|duplicate|iteration\\s+assignment|field\\s+classification"

  - name: instruction-ado-create-pull-request-conformance
    prompt: |
      You are creating an Azure DevOps pull request from artifacts under
      `.copilot-tracking/pr/new/feature-x/`. Which
      `.github/instructions/**/*.instructions.md` file applies (cite its path)
      and what does it require for work item discovery and reviewer
      identification?
    tags:
      category: behavior-conformance
      instruction: ado-create-pull-request
      advisory: "true"
    graders:
      - type: output-matches
        name: applyTo-evidence
        config:
          pattern: "(?i)\\.copilot-tracking/pr/new|pr/new/"
      - type: output-matches
        name: scope-language
        config:
          pattern: "(?i)work\\s+item|reviewer|pull\\s+request|automated\\s+linking"

  - name: instruction-ado-get-build-info-conformance
    prompt: |
      A user asks about Azure DevOps build status for a branch and you are
      drafting a response file at `.copilot-tracking/pr/123-build-info.md`.
      Which `.github/instructions/**/*.instructions.md` file applies (cite
      its path) and what does it require?
    tags:
      category: behavior-conformance
      instruction: ado-get-build-info
      advisory: "true"
    graders:
      - type: output-matches
        name: applyTo-evidence
        config:
          pattern: "(?i)\\.copilot-tracking/pr/.*build|pr/.*-build-"
      - type: output-matches
        name: scope-language
        config:
          pattern: "(?i)azure\\s*devops|\\bado\\b|build|pull\\s*request|branch"

  - name: instruction-ado-update-wit-items-conformance
    prompt: |
      You are about to execute work item create/update operations from a
      handoff log at `.copilot-tracking/workitems/sprint-12/handoff-logs.md`.
      Which `.github/instructions/**/*.instructions.md` file applies (cite
      its path) and what does it require about MCP ADO tools and handoff
      tracking?
    tags:
      category: behavior-conformance
      instruction: ado-update-wit-items
      advisory: "true"
    graders:
      - type: output-matches
        name: applyTo-evidence
        config:
          pattern: "(?i)handoff-logs\\.md|workitems/.*handoff"
      - type: output-matches
        name: scope-language
        config:
          pattern: "(?i)mcp|ado|work\\s+item|handoff"

  - name: instruction-ado-wit-discovery-conformance
    prompt: |
      You are running ADO work item discovery and writing output under
      `.copilot-tracking/workitems/discovery/team-roadmap/`. Which
      `.github/instructions/**/*.instructions.md` file applies (cite its
      path) and what does it require?
    tags:
      category: behavior-conformance
      instruction: ado-wit-discovery
      advisory: "true"
    graders:
      - type: output-matches
        name: applyTo-evidence
        config:
          pattern: "(?i)\\.copilot-tracking/workitems/discovery|workitems/discovery/"
      - type: output-matches
        name: scope-language
        config:
          pattern: "(?i)discover|assignment|artifact|planning\\s+file"

  - name: instruction-ado-wit-planning-conformance
    prompt: |
      You are writing an ADO work item planning file under
      `.copilot-tracking/workitems/prd-payments/issues-plan.md`. Which
      `.github/instructions/**/*.instructions.md` file is the reference
      specification (cite its path) and what does it require for templates,
      field definitions, and search protocols?
    tags:
      category: behavior-conformance
      instruction: ado-wit-planning
      advisory: "true"
    graders:
      - type: output-matches
        name: applyTo-evidence
        config:
          pattern: "(?i)\\.copilot-tracking/workitems|workitems/"
      - type: output-matches
        name: scope-language
        config:
          pattern: "(?i)template|field|search\\s+protocol|planning\\s+file"

  - name: instruction-docusaurus-edits-conformance
    prompt: |
      You are creating a new Docusaurus documentation page at
      `docs/rpi/new-page.md`. Which `.github/instructions/**/*.instructions.md`
      file applies (cite its path) and what conventions does it require for
      frontmatter, admonitions, and internal links?
    tags:
      category: behavior-conformance
      instruction: docusaurus-edits
      advisory: "true"
    graders:
      - type: output-matches
        name: applyTo-evidence
        config:
          pattern: "(?i)\\bdocs/\\b|^docs/|docs\\s*/"
      - type: output-matches
        name: scope-language
        config:
          pattern: "(?i)docusaurus|frontmatter|admonition|internal\\s+link|mermaid"

  - name: instruction-experiment-designer-conformance
    prompt: |
      You are working on a Minimum Viable Experimentation session under
      `.copilot-tracking/mve/payments-latency/`. Which
      `.github/instructions/**/*.instructions.md` file applies (cite its
      path) and what MVE coaching conventions does it require?
    tags:
      category: behavior-conformance
      instruction: experiment-designer
      advisory: "true"
    graders:
      - type: output-matches
        name: applyTo-evidence
        config:
          pattern: "(?i)\\.copilot-tracking/mve|/mve/"
      - type: output-matches
        name: scope-language
        config:
          pattern: "(?i)\\bmve\\b|experiment|hypothes|assumption|validate"

  - name: instruction-github-backlog-discovery-conformance
    prompt: |
      You are running GitHub issue discovery and writing output under
      `.copilot-tracking/github-issues/discovery/v2-features/`. Which
      `.github/instructions/**/*.instructions.md` file applies (cite its
      path) and what discovery paths does it require?
    tags:
      category: behavior-conformance
      instruction: github-backlog-discovery
      advisory: "true"
    graders:
      - type: output-matches
        name: applyTo-evidence
        config:
          pattern: "(?i)\\.copilot-tracking/github-issues/discovery|github-issues/discovery"
      - type: output-matches
        name: scope-language
        config:
          pattern: "(?i)discover|user-centric|artifact-driven|search-based|github\\s+issue"

  - name: instruction-github-backlog-planning-conformance
    prompt: |
      You are creating GitHub backlog planning files under
      `.copilot-tracking/github-issues/v3-roadmap/`. Which
      `.github/instructions/**/*.instructions.md` file is the reference
      specification (cite its path) and what does it require for templates,
      search protocols, and state persistence?
    tags:
      category: behavior-conformance
      instruction: github-backlog-planning
      advisory: "true"
    graders:
      - type: output-matches
        name: applyTo-evidence
        config:
          pattern: "(?i)\\.copilot-tracking/github-issues|github-issues/"
      - type: output-matches
        name: scope-language
        config:
          pattern: "(?i)template|search\\s+protocol|state\\s+persistence|similarity"

  - name: instruction-github-backlog-triage-conformance
    prompt: |
      You are triaging untriaged GitHub issues with artifacts under
      `.copilot-tracking/github-issues/triage/q1-cleanup/`. Which
      `.github/instructions/**/*.instructions.md` file applies (cite its
      path) and what does it require for label suggestion, milestone
      assignment, and duplicate detection?
    tags:
      category: behavior-conformance
      instruction: github-backlog-triage
      advisory: "true"
    graders:
      - type: output-matches
        name: applyTo-evidence
        config:
          pattern: "(?i)\\.copilot-tracking/github-issues/triage|github-issues/triage"
      - type: output-matches
        name: scope-language
        config:
          pattern: "(?i)triage|label|milestone|duplicate|conventional\\s+commit"

  - name: instruction-github-backlog-update-conformance
    prompt: |
      You are about to execute GitHub issue operations from a handoff log at
      `.copilot-tracking/github-issues/v2-features/handoff-logs.md`. Which
      `.github/instructions/**/*.instructions.md` file applies (cite its
      path) and what does the execution workflow require?
    tags:
      category: behavior-conformance
      instruction: github-backlog-update
      advisory: "true"
    graders:
      - type: output-matches
        name: applyTo-evidence
        config:
          pattern: "(?i)github-issues/.*handoff-logs\\.md|handoff-logs\\.md"
      - type: output-matches
        name: scope-language
        config:
          pattern: "(?i)handoff|sequential|create|update|link|close|github"

  - name: instruction-markdown-conformance
    prompt: |
      You are creating or editing a Markdown file at `docs/example.md`. Which
      `.github/instructions/**/*.instructions.md` file defines the style
      guide that applies (cite its path) and what are its top three rules
      for headings and list formatting?
    tags:
      category: behavior-conformance
      instruction: markdown
      advisory: "true"
    graders:
      - type: output-matches
        name: applyTo-evidence
        config:
          pattern: "(?i)\\*\\*/\\*\\.md|\\.md\\b|all\\s+markdown\\s+files"
      - type: output-matches
        name: scope-language
        config:
          pattern: "(?i)markdownlint|heading|atx|list|frontmatter"

  - name: instruction-prompt-builder-conformance
    prompt: |
      You are authoring a new custom agent file at
      `.github/agents/example/my-agent.agent.md`. Which
      `.github/instructions/**/*.instructions.md` file applies (cite its
      path) and what authoring standards does it require for prompt
      engineering artifacts?
    tags:
      category: behavior-conformance
      instruction: prompt-builder
      advisory: "true"
    graders:
      - type: output-matches
        name: applyTo-evidence
        config:
          pattern: "(?i)\\.agent\\.md|\\.instructions\\.md|\\.prompt\\.md|SKILL\\.md"
      - type: output-matches
        name: scope-language
        config:
          pattern: "(?i)prompt|agent|instruction|skill|frontmatter|authoring"

  - name: instruction-pull-request-conformance
    prompt: |
      You are generating a pull request description from branch diffs and
      writing tracking artifacts under `.copilot-tracking/pr/main-feature/`.
      Which `.github/instructions/**/*.instructions.md` file applies (cite
      its path) and what does it require for diff analysis and template
      handling?
    tags:
      category: behavior-conformance
      instruction: pull-request
      advisory: "true"
    graders:
      - type: output-matches
        name: applyTo-evidence
        config:
          pattern: "(?i)\\.copilot-tracking/pr/|/pr/"
      - type: output-matches
        name: scope-language
        config:
          pattern: "(?i)pull\\s+request|diff|template|checkbox|pr-reference"

  - name: instruction-writing-style-conformance
    prompt: |
      You are writing prose in a Markdown documentation file at
      `docs/getting-started/install.md`. Which
      `.github/instructions/**/*.instructions.md` file defines the writing
      style conventions that apply (cite its path) and what are its core
      voice and tone requirements?
    tags:
      category: behavior-conformance
      instruction: writing-style
      advisory: "true"
    graders:
      - type: output-matches
        name: applyTo-evidence
        config:
          pattern: "(?i)\\*\\*/\\*\\.md|\\.md\\b|markdown\\s+content"
      - type: output-matches
        name: scope-language
        config:
          pattern: "(?i)voice|tone|style|formal|instructional|professional"

  - name: instruction-jira-backlog-planning-conformance
    prompt: |
      You are creating Jira backlog planning files under
      `.copilot-tracking/jira-issues/migration-plan/`. Which
      `.github/instructions/**/*.instructions.md` file is the reference
      specification (cite its path) and what conventions does it require for
      planning, search, and state persistence?
    tags:
      category: behavior-conformance
      instruction: jira-backlog-planning
      advisory: "true"
    graders:
      - type: output-matches
        name: applyTo-evidence
        config:
          pattern: "(?i)\\.copilot-tracking/jira-issues|jira-issues/"
      - type: output-matches
        name: scope-language
        config:
          pattern: "(?i)\\bjira\\b|template|jql|search|state\\s+persistence"

  - name: instruction-jira-wit-planning-conformance
    prompt: |
      You are working on Jira PRD-driven work item planning files under
      `.copilot-tracking/jira-issues/prds/payments-prd/`. Which
      `.github/instructions/**/*.instructions.md` file is the reference
      specification (cite its path) and what does it require for hierarchy
      mapping and field validation?
    tags:
      category: behavior-conformance
      instruction: jira-wit-planning
      advisory: "true"
    graders:
      - type: output-matches
        name: applyTo-evidence
        config:
          pattern: "(?i)\\.copilot-tracking/jira-issues/prds|jira-issues/prds"
      - type: output-matches
        name: scope-language
        config:
          pattern: "(?i)\\bprd\\b|hierarchy|field|handoff|issue\\s+type"

  - name: instruction-rai-identity-conformance
    prompt: |
      You are running an RAI planning session with artifacts under
      `.copilot-tracking/rai-plans/customer-onboarding/`. Which
      `.github/instructions/**/*.instructions.md` file defines the RAI
      Planner identity and orchestration (cite its path) and what six-phase
      workflow does it require?
    tags:
      category: behavior-conformance
      instruction: rai-identity
      advisory: "true"
    graders:
      - type: output-matches
        name: applyTo-evidence
        config:
          pattern: "(?i)\\.copilot-tracking/rai-plans|rai-plans/"
      - type: output-matches
        name: scope-language
        config:
          pattern: "(?i)rai|nist|six[-\\s]phase|trustworthi|assessment"

  - name: instruction-rai-risk-classification-conformance
    prompt: |
      You are completing Phase 2 risk classification for an RAI session at
      `.copilot-tracking/rai-plans/credit-scoring/`. Which
      `.github/instructions/**/*.instructions.md` file applies (cite its
      path) and what does it require for the prohibited uses gate and depth
      tier assignment?
    tags:
      category: behavior-conformance
      instruction: rai-identity
      advisory: "true"
    graders:
      - type: output-matches
        name: applyTo-evidence
        config:
          pattern: "(?i)\\.copilot-tracking/rai-plans|rai-plans/"
      - type: output-matches
        name: scope-language
        config:
          pattern: "(?i)prohibited\\s+use|risk\\s+(indicator|classification)|depth\\s+tier|basic|standard|comprehensive"

  - name: instruction-backlog-handoff-conformance
    prompt: |
      You are completing Phase 5 of a Security Planning session and
      generating backlog work items under
      `.copilot-tracking/security-plans/order-service/`. Which
      `.github/instructions/**/*.instructions.md` file applies (cite its
      path) and what dual-format ADO and GitHub work item templates does it
      require?
    tags:
      category: behavior-conformance
      instruction: identity
      advisory: "true"
    graders:
      - type: output-matches
        name: applyTo-evidence
        config:
          pattern: "(?i)\\.copilot-tracking/security-plans|security-plans/"
      - type: output-matches
        name: scope-language
        config:
          pattern: "(?i)backlog|handoff|ado|github|wi-sec|work\\s+item|stride|mitigation"

  - name: instruction-sssc-assessment-conformance
    prompt: |
      You are running an SSSC Phase 2 supply chain assessment with artifacts
      under `.copilot-tracking/sssc-plans/payments-toolchain/`. Which
      `.github/instructions/**/*.instructions.md` file applies (cite its
      path) and what does it require for the combined capabilities inventory?
    tags:
      category: behavior-conformance
      instruction: sssc-planner
      advisory: "true"
    graders:
      - type: output-matches
        name: applyTo-evidence
        config:
          pattern: "(?i)\\.copilot-tracking/sssc-plans|sssc-plans/"
      - type: output-matches
        name: scope-language
        config:
          pattern: "(?i)sssc|supply\\s+chain|capabilit|sbom|scorecard|slsa|sigstore"

  - name: instruction-story-quality-conformance
    prompt: |
      You are authoring a new ADO User Story or refining one in a custom
      agent file at `.github/agents/ado/my-backlog-agent.agent.md`. Which
      `.github/instructions/**/*.instructions.md` file defines the shared
      story quality conventions (cite its path) and what does it require for
      titles, descriptions, and acceptance criteria?
    tags:
      category: behavior-conformance
      instruction: story-quality
      advisory: "true"
    graders:
      - type: output-matches
        name: applyTo-evidence
        config:
          pattern: "(?i)\\.agent\\.md|\\.github/instructions/ado/"
      - type: output-matches
        name: scope-language
        config:
          pattern: "(?i)story|title|description|acceptance\\s+criteria|user\\s+story|goal\\s+statement|problem\\s+statement"

  - name: instruction-disclaimer-language-conformance
    prompt: |
      You are starting an RAI planning session and need to display the
      required disclaimer at startup, with artifacts under
      `.copilot-tracking/rai-plans/new-project/`. Which
      `.github/instructions/**/*.instructions.md` file centralizes the
      disclaimer language (cite its path) and what does it require?
    tags:
      category: behavior-conformance
      instruction: disclaimer-language
      advisory: "true"
    graders:
      - type: output-matches
        name: applyTo-evidence
        config:
          pattern: "(?i)\\.copilot-tracking/(rai-plans|security-plans|sssc-plans)"
      - type: output-matches
        name: scope-language
        config:
          pattern: "(?i)disclaimer|caution|assistive\\s+tool|professional\\s+review|legal|compliance"

  - name: instruction-powershell-script-conformance
    prompt: |
      You are about to author a new PowerShell automation script at
      `scripts/automation/Invoke-Deploy.ps1`. Which
      `.github/instructions/**/*.instructions.md` file applies (cite its
      path) and what does it require for the shebang, copyright header,
      `[CmdletBinding()]` param block, and `$ErrorActionPreference`?
    tags:
      category: behavior-conformance
      instruction: powershell
      advisory: "true"
    graders:
      - type: output-matches
        name: applyTo-evidence
        config:
          pattern: "(?i)\\*\\.ps1|\\*\\.psm1|\\*\\.psd1|powershell\\.instructions\\.md"
      - type: output-matches
        name: scope-language
        config:
          pattern: "(?i)cmdletbinding|param\\b|erroractionpreference|comment-based\\s+help|synopsis|\\.ps1|pwsh"

  - name: instruction-powershell-module-conformance
    prompt: |
      You are authoring a new PowerShell module at
      `scripts/lib/Modules/DeploymentHelpers.psm1`. Which
      `.github/instructions/**/*.instructions.md` file applies (cite its
      path) and how does the module structure differ from a `.ps1` script
      (purpose comment, `#Requires`, `Export-ModuleMember`)?
    tags:
      category: behavior-conformance
      instruction: powershell
      advisory: "true"
    graders:
      - type: output-matches
        name: applyTo-evidence
        config:
          pattern: "(?i)\\*\\.psm1|\\*\\.ps1|powershell\\.instructions\\.md"
      - type: output-matches
        name: scope-language
        config:
          pattern: "(?i)export-modulemember|#requires|module|psm1|outputtype|using\\s+module"

  - name: instruction-powershell-data-file-conformance
    prompt: |
      You are creating a PowerShell data file (module manifest) at
      `scripts/lib/Modules/CIHelpers.psd1`. Which
      `.github/instructions/**/*.instructions.md` file applies (cite its
      path) and what does it cover for `.psd1` files?
    tags:
      category: behavior-conformance
      instruction: powershell
      advisory: "true"
    graders:
      - type: output-matches
        name: applyTo-evidence
        config:
          pattern: "(?i)\\*\\.psd1|\\*\\.psm1|\\*\\.ps1|powershell\\.instructions\\.md"
      - type: output-matches
        name: scope-language
        config:
          pattern: "(?i)\\.psd1|manifest|module|powershell"

  - name: instruction-powershell-comment-help-conformance
    prompt: |
      You are adding comment-based help to an existing PowerShell script at
      `scripts/linting/Invoke-Validation.ps1`. Which
      `.github/instructions/**/*.instructions.md` file applies (cite its
      path) and which help keywords are required (SYNOPSIS, DESCRIPTION,
      PARAMETER, EXAMPLE, NOTES)?
    tags:
      category: behavior-conformance
      instruction: powershell
      advisory: "true"
    graders:
      - type: output-matches
        name: applyTo-evidence
        config:
          pattern: "(?i)\\*\\.ps1|powershell\\.instructions\\.md"
      - type: output-matches
        name: scope-language
        config:
          pattern: "(?i)\\.synopsis|\\.description|\\.parameter|\\.example|\\.notes|comment-based\\s+help"

  - name: instruction-powershell-error-handling-conformance
    prompt: |
      You are adding error handling and exit codes to a PowerShell script at
      `scripts/security/Test-Permissions.ps1`. Which
      `.github/instructions/**/*.instructions.md` file applies (cite its
      path) and what does it require for `$ErrorActionPreference`, try/catch
      placement, `$LASTEXITCODE` checks, and the main execution guard?
    tags:
      category: behavior-conformance
      instruction: powershell
      advisory: "true"
    graders:
      - type: output-matches
        name: applyTo-evidence
        config:
          pattern: "(?i)\\*\\.ps1|powershell\\.instructions\\.md"
      - type: output-matches
        name: scope-language
        config:
          pattern: "(?i)erroractionpreference|try.*catch|lastexitcode|throw|invocation(name)?|exit\\s+1"

  - name: instruction-python-script-cli-conformance
    prompt: |
      You are authoring a new Python CLI script at
      `scripts/automation/process_files.py` that uses argparse. Which
      `.github/instructions/**/*.instructions.md` file applies (cite its
      path) and what does it require for argument parsing, type hints, and
      `main()` entry-point structure?
    tags:
      category: behavior-conformance
      instruction: python-script
      advisory: "true"
    graders:
      - type: output-matches
        name: applyTo-evidence
        config:
          pattern: "(?i)\\*\\.py|python-script\\.instructions\\.md"
      - type: output-matches
        name: scope-language
        config:
          pattern: "(?i)argparse|click|create_parser|def\\s+main|sys\\.exit|exit\\s+code|type\\s+hint"

  - name: instruction-python-script-pathlib-conformance
    prompt: |
      You are updating a Python automation script at
      `scripts/data/transform_inputs.py` to handle file paths. Which
      `.github/instructions/**/*.instructions.md` file applies (cite its
      path) and what does it say about `pathlib.Path` versus `os.path`?
    tags:
      category: behavior-conformance
      instruction: python-script
      advisory: "true"
    graders:
      - type: output-matches
        name: applyTo-evidence
        config:
          pattern: "(?i)\\*\\.py|python-script\\.instructions\\.md"
      - type: output-matches
        name: scope-language
        config:
          pattern: "(?i)pathlib|\\bPath\\b|read_text|write_text|os\\.path"

  - name: instruction-python-script-subprocess-conformance
    prompt: |
      You are writing a Python helper at `scripts/automation/run_linter.py`
      that shells out to other tools. Which
      `.github/instructions/**/*.instructions.md` file applies (cite its
      path) and what does it require for `subprocess.run`, `capture_output`,
      `check=True`, and error handling?
    tags:
      category: behavior-conformance
      instruction: python-script
      advisory: "true"
    graders:
      - type: output-matches
        name: applyTo-evidence
        config:
          pattern: "(?i)\\*\\.py|python-script\\.instructions\\.md"
      - type: output-matches
        name: scope-language
        config:
          pattern: "(?i)subprocess|capture_output|check=true|calledprocesserror|filenotfounderror"

  - name: instruction-python-tests-pytest-conformance
    prompt: |
      You are adding a new pytest test file at
      `scripts/tests/test_validator.py`. Which
      `.github/instructions/**/*.instructions.md` file applies (cite its
      path) and what does it require for test naming, AAA structure,
      mocking, and parametrize usage?
    tags:
      category: behavior-conformance
      instruction: python-tests
      advisory: "true"
    graders:
      - type: output-matches
        name: applyTo-evidence
        config:
          pattern: "(?i)\\*\\.py|python-tests\\.instructions\\.md"
      - type: output-matches
        name: scope-language
        config:
          pattern: "(?i)pytest|mocker|monkeypatch|arrange|act|assert|parametrize|test_given_"

  - name: instruction-python-uv-projects-conformance
    prompt: |
      You are setting up a new Python skill virtual environment under
      `.github/skills/data-science/example-skill/` and need to manage
      dependencies with uv. Which `.github/instructions/**/*.instructions.md`
      file applies (cite its path) and which uv commands does it prescribe
      for init, venv creation, dependency add, sync, and lock?
    tags:
      category: behavior-conformance
      instruction: uv-projects
      advisory: "true"
    graders:
      - type: output-matches
        name: applyTo-evidence
        config:
          pattern: "(?i)\\*\\.py|\\*\\.ipynb|uv-projects\\.instructions\\.md"
      - type: output-matches
        name: scope-language
        config:
          pattern: "(?i)\\buv\\s+(init|venv|add|sync|lock)|pyproject\\.toml|\\.venv|ipykernel|ruff"

  - name: instruction-rust-cargo-conformance
    prompt: |
      You are creating a new Rust service crate with `Cargo.toml` at
      `crates/order-service/Cargo.toml` plus `src/main.rs`. Which
      `.github/instructions/**/*.instructions.md` file applies (cite its
      path) and what does it require for `[package]` metadata, edition,
      dependency pinning, and the release profile?
    tags:
      category: behavior-conformance
      instruction: rust
      advisory: "true"
    graders:
      - type: output-matches
        name: applyTo-evidence
        config:
          pattern: "(?i)\\*\\.rs|rust\\.instructions\\.md"
      - type: output-matches
        name: scope-language
        config:
          pattern: "(?i)cargo\\.toml|edition\\s*=\\s*\"?2021|\\[package\\]|\\[profile\\.release\\]|lto|codegen-units"

  - name: instruction-rust-error-handling-conformance
    prompt: |
      You are adding error types to a Rust library crate at
      `crates/order-service/src/error.rs`. Which
      `.github/instructions/**/*.instructions.md` file applies (cite its
      path) and what does it require for `thiserror`, the `Result<T>` alias,
      `#[from]` delegation, and avoidance of `unwrap()` in production paths?
    tags:
      category: behavior-conformance
      instruction: rust
      advisory: "true"
    graders:
      - type: output-matches
        name: applyTo-evidence
        config:
          pattern: "(?i)\\*\\.rs|rust\\.instructions\\.md"
      - type: output-matches
        name: scope-language
        config:
          pattern: "(?i)thiserror|anyhow|#\\[from\\]|unwrap|expect|Result<|ServiceError|\\?\\s+operator|error\\s+propagation"

  - name: instruction-rust-naming-conformance
    prompt: |
      You are naming new types, traits, functions, and constants in a Rust
      module at `crates/order-service/src/services.rs`. Which
      `.github/instructions/**/*.instructions.md` file applies (cite its
      path) and what naming conventions and type-suffix conventions does it
      require (PascalCase, snake_case, SCREAMING_SNAKE_CASE, `*Error`,
      `*Config`, `*Builder`)?
    tags:
      category: behavior-conformance
      instruction: rust
      advisory: "true"
    graders:
      - type: output-matches
        name: applyTo-evidence
        config:
          pattern: "(?i)\\*\\.rs|rust\\.instructions\\.md"
      - type: output-matches
        name: scope-language
        config:
          pattern: "(?i)pascalcase|snake_case|screaming_snake_case|kebab-case|\\*Error|\\*Config|\\*Builder"

  - name: instruction-rust-async-conformance
    prompt: |
      You are adding async functions backed by Tokio to a Rust service at
      `crates/order-service/src/main.rs`. Which
      `.github/instructions/**/*.instructions.md` file applies (cite its
      path) and what does it require for Tokio runtime flavor selection and
      `#[tokio::main]` usage?
    tags:
      category: behavior-conformance
      instruction: rust
      advisory: "true"
    graders:
      - type: output-matches
        name: applyTo-evidence
        config:
          pattern: "(?i)\\*\\.rs|rust\\.instructions\\.md"
      - type: output-matches
        name: scope-language
        config:
          pattern: "(?i)tokio|#\\[tokio::main\\]|async\\s+fn|multi-threaded|runtime"

  - name: instruction-rust-tests-conformance
    prompt: |
      You are writing Rust unit tests inside
      `crates/order-service/src/services.rs` and integration tests under
      `crates/order-service/tests/integration_test.rs`. Which
      `.github/instructions/**/*.instructions.md` file applies (cite its
      path) and what does it require for `#[cfg(test)]` module placement,
      `#[tokio::test]`, and test naming?
    tags:
      category: behavior-conformance
      instruction: rust-tests
      advisory: "true"
    graders:
      - type: output-matches
        name: applyTo-evidence
        config:
          pattern: "(?i)\\*\\.rs|rust-tests\\.instructions\\.md"
      - type: output-matches
        name: scope-language
        config:
          pattern: "(?i)#\\[cfg\\(test\\)\\]|#\\[test\\]|#\\[tokio::test\\]|mod\\s+tests|given_.*_when_.*_then_|integration"

  - name: instruction-bash-script-structure-conformance
    prompt: |
      You are creating a new bash automation script at
      `scripts/install/configure-host.sh`. Which
      `.github/instructions/**/*.instructions.md` file applies (cite its
      path) and what does it require for the shebang, copyright header,
      strict mode, and `main()` function pattern?
    tags:
      category: behavior-conformance
      instruction: bash
      advisory: "true"
    graders:
      - type: output-matches
        name: applyTo-evidence
        config:
          pattern: "(?i)\\*\\.sh|bash\\.instructions\\.md"
      - type: output-matches
        name: scope-language
        config:
          pattern: "(?i)#!/usr/bin/env\\s+bash|set\\s+-euo\\s+pipefail|main\\(\\)|copyright|shellcheck"

  - name: instruction-bash-naming-conformance
    prompt: |
      You are reviewing variable and function names in a bash script at
      `scripts/deploy/deploy-app.sh`. Which
      `.github/instructions/**/*.instructions.md` file applies (cite its
      path) and what naming conventions does it require for environment
      variables, constants, local variables, and functions?
    tags:
      category: behavior-conformance
      instruction: bash
      advisory: "true"
    graders:
      - type: output-matches
        name: applyTo-evidence
        config:
          pattern: "(?i)\\*\\.sh|bash\\.instructions\\.md"
      - type: output-matches
        name: scope-language
        config:
          pattern: "(?i)upper_snake_case|lower_snake_case|readonly|\\blocal\\b|variable\\s+expansion|\\$\\{|braces"

  - name: instruction-bash-error-handling-conformance
    prompt: |
      You are adding error handling to a bash script at
      `scripts/maintenance/cleanup-logs.sh`. Which
      `.github/instructions/**/*.instructions.md` file applies (cite its
      path) and what does it require for `set -euo pipefail`, `err()`
      functions, and required variable checks?
    tags:
      category: behavior-conformance
      instruction: bash
      advisory: "true"
    graders:
      - type: output-matches
        name: applyTo-evidence
        config:
          pattern: "(?i)\\*\\.sh|bash\\.instructions\\.md"
      - type: output-matches
        name: scope-language
        config:
          pattern: "(?i)set\\s+-euo\\s+pipefail|\\berr\\(\\)|exit\\s+1|stderr|>&2|shellcheck"

  - name: instruction-pptx-conformance
    prompt: |
      You are building a PowerPoint deck and authoring content under
      `.copilot-tracking/ppt/quarterly-review/`. Which
      `.github/instructions/**/*.instructions.md` file applies (cite its
      path) and what conventions does it require for the agent, subagent,
      and powerpoint skill workflows?
    tags:
      category: behavior-conformance
      instruction: pptx
      advisory: "true"
    graders:
      - type: output-matches
        name: applyTo-evidence
        config:
          pattern: "(?i)\\.copilot-tracking/ppt/|/ppt/|pptx\\.instructions\\.md"
      - type: output-matches
        name: scope-language
        config:
          pattern: "(?i)powerpoint|slide|deck|design\\s+rule|convention|skill"