Contract
- Input: file-save event, changed file paths, project tooling config.
- Output:
GateIdeArtifact with per-check results and terminal state.
- Scope: IDE-level checks only — lint, format, typecheck, spellcheck, secret scan.
- Rule: run checks in parallel where possible to stay under sub-second budget.
- Rule: lightweight secret scan only — do not run full dependency audit here.
- Rule: report results to agent context immediately; do not block the save operation itself.
Provenance
| Question |
Answer |
| What is the source of truth? |
The file-save event and project tooling config (eslint, prettier, typescript, cspell, gitleaks) |
| What is in scope? |
Changed files since last save; IDE-level checks only |
| What is explicitly out of scope? |
Full test suite, dependency audit, security scan, performance tests |
| Who or what consumes this artifact afterward? |
gate-pre-commit or agent context for immediate feedback |
| What evidence proves it is done? |
All checks completed, results reported, structured result returned |
| What risk remains? |
False positives from spellchecker; secret scan is lightweight and may miss obfuscated secrets |
Artifact
Return GateIdeArtifact as structured output in the current response. Do not write local artifact files unless the user explicitly asks for an export.
Process
1. Detect changed files
Receive the file-save event payload. Extract the list of changed file paths. If no files changed, skip checks and report idle.
2. Run lint
Execute the project's linter (eslint, pylint, rubocop, etc.) against changed files only. Capture exit code, stdout, and stderr.
3. Run format check
Execute the formatter in check mode (prettier --check, black --check, etc.). Do not modify files.
4. Run typecheck
Execute the type checker (tsc --noEmit, mypy, etc.) against changed files and their dependents.
5. Run spellcheck
Execute spellchecker (cspell, codespell) against changed files. Flag unknown words but do not block on them unless configured as errors.
6. Run lightweight secret scan
Scan changed files for common secret patterns (API keys, tokens, passwords). Use regex-based detection only — do not run full gitleaks or truffleHog here.
7. Aggregate results
Compile all check results into GateIdeArtifact. Determine terminal state:
- If any check failed and
blockOnFailure is true → blocked
- If all checks passed →
completed
- If checks errored →
failed
8. Report to agent context
Publish a concise summary to agent context so the user sees results immediately.
Completion
- all 5 checks have run or been explicitly skipped with reason
- structured result returned
- results reported to agent context
- terminal state correctly reflects
blockOnFailure policy
1---2name: gate-ide3category: platform4maturity: stable5version: 16description: Shift-left quality gate triggered on file-save events. Runs lint, format, typecheck, spellcheck, and lightweight secret scan. Blocks on failure and reports results to agent context.7license: MIT8capabilities:9 - run ide-level checks10 - produce quality gate artifact11 - validate quality gate completion criteria12 - trace gate ide decisions across inputs, constraints, and failure paths13 - validate platform outcomes with evidence, assumptions, and explicit next steps14outputs:15 - type: object16 name: GateIdeArtifact17 properties:18 trigger: { type: string, enum: [file-save] }19 checks:20 type: array21 items:22 type: object23 properties:24 name: { type: string }25 status: { type: string, enum: [passed, failed, skipped] }26 durationMs: { type: integer }27 output: { type: string }28 blockOnFailure: { type: boolean }29 reportTo: { type: string }30 result: { type: string, enum: [completed, blocked, failed] }31 required: [trigger, checks, blockOnFailure, reportTo, result]32sideEffects: []33dependencies: []34stopCondition: Shift-left quality gate triggered on file-save events complete; structured result returned; completion criteria checked.35risk: low36trustTier: 137maxIterations: 338modelTier: fast39promptVersion: "2.0"40artifactType: plan41evaluators:42 - behavioral43 - regression44fixturesPath: .agents/skills/platform/fixtures/behavioral/gate-ide.json45diataxis: how-to46tags: [quality, shift-left, ide, gate]47compatibility: [implement, tdd, publish-open-pr]48approvalRequired: false49approvalFor: []50metadata:51 source: https://skillmd.com/skills/quantumquirkxyz/gate-ide52---5354## Contract5556- Input: file-save event, changed file paths, project tooling config.57- Output: `GateIdeArtifact` with per-check results and terminal state.58- Scope: IDE-level checks only — lint, format, typecheck, spellcheck, secret scan.59- Rule: run checks in parallel where possible to stay under sub-second budget.60- Rule: lightweight secret scan only — do not run full dependency audit here.61- Rule: report results to agent context immediately; do not block the save operation itself.6263## Provenance6465| Question | Answer |66|---|---|67| What is the source of truth? | The file-save event and project tooling config (eslint, prettier, typescript, cspell, gitleaks) |68| What is in scope? | Changed files since last save; IDE-level checks only |69| What is explicitly out of scope? | Full test suite, dependency audit, security scan, performance tests |70| Who or what consumes this artifact afterward? | `gate-pre-commit` or agent context for immediate feedback |71| What evidence proves it is done? | All checks completed, results reported, structured result returned |72| What risk remains? | False positives from spellchecker; secret scan is lightweight and may miss obfuscated secrets |7374## Artifact7576Return `GateIdeArtifact` as structured output in the current response. Do not write local artifact files unless the user explicitly asks for an export.7778## Process7980### 1. Detect changed files81Receive the file-save event payload. Extract the list of changed file paths. If no files changed, skip checks and report idle.8283### 2. Run lint84Execute the project's linter (eslint, pylint, rubocop, etc.) against changed files only. Capture exit code, stdout, and stderr.8586### 3. Run format check87Execute the formatter in check mode (prettier --check, black --check, etc.). Do not modify files.8889### 4. Run typecheck90Execute the type checker (tsc --noEmit, mypy, etc.) against changed files and their dependents.9192### 5. Run spellcheck93Execute spellchecker (cspell, codespell) against changed files. Flag unknown words but do not block on them unless configured as errors.9495### 6. Run lightweight secret scan96Scan changed files for common secret patterns (API keys, tokens, passwords). Use regex-based detection only — do not run full gitleaks or truffleHog here.9798### 7. Aggregate results99Compile all check results into `GateIdeArtifact`. Determine terminal state:100- If any check failed and `blockOnFailure` is true → `blocked`101- If all checks passed → `completed`102- If checks errored → `failed`103104### 8. Report to agent context105Publish a concise summary to agent context so the user sees results immediately.106107## Completion108109- all 5 checks have run or been explicitly skipped with reason110- structured result returned111- results reported to agent context112- terminal state correctly reflects `blockOnFailure` policy