grade-on-ingest
Trigger GRADE quality assessment automatically when new research sources or findings enter the corpus
Best use case
grade-on-ingest is best used when you need a repeatable AI agent workflow instead of a one-off prompt.
It is a strong fit for teams already working in Codex.
Trigger GRADE quality assessment automatically when new research sources or findings enter the corpus
Teams using grade-on-ingest should expect a more consistent output, faster repeated execution, less prompt rewriting.
When to use this skill
- You want a reusable workflow that can be run more than once with consistent structure.
When not to use this skill
- You only need a quick one-off answer and do not need a reusable workflow.
- You cannot install or maintain the underlying files, dependencies, or repository context.
Installation
Claude Code / Cursor / Codex
Manual Installation
- Download SKILL.md from GitHub
- Place it in
.claude/skills/grade-on-ingest/SKILL.mdinside your project - Restart your AI agent — it will auto-discover the skill
How grade-on-ingest Compares
| Feature / Agent | grade-on-ingest | Standard Approach |
|---|---|---|
| Platform Support | Codex | Limited / Varies |
| Context Awareness | High | Baseline |
| Installation Complexity | Unknown | N/A |
Frequently Asked Questions
What does this skill do?
Trigger GRADE quality assessment automatically when new research sources or findings enter the corpus
Which AI agents support this skill?
This skill is designed for Codex.
Where can I find the source code?
You can find the source code on GitHub using the link provided at the top of the page.
Related Guides
AI Agents for Coding
Browse AI agent skills for coding, debugging, testing, refactoring, code review, and developer workflows across Claude, Cursor, and Codex.
Cursor vs Codex for AI Workflows
Compare Cursor and Codex for AI coding workflows, repository assistance, debugging, refactoring, and reusable developer skills.
AI Agents for Startups
Explore AI agent skills for startup validation, product research, growth experiments, documentation, and fast execution with small teams.
SKILL.md Source
# GRADE-on-Ingest
Automatically triggers GRADE quality assessment when new research sources or findings are added to the corpus.
## Triggers
Alternate expressions and non-obvious activations (primary phrases are matched automatically from the skill description):
- "GRADE" → evidence quality rating framework
- "quality of evidence" → GRADE assessment
- "evidence level" → source quality grading
## Purpose
Ensures every research source entering the corpus receives a GRADE quality assessment at ingestion time, preventing unassessed sources from being cited without quality context. Implements the "assess at entry" pattern to maintain corpus-wide quality visibility.
## Activation Conditions
This skill activates when:
1. **New file created** in `.aiwg/research/sources/` or `.aiwg/research/findings/`
2. **File pattern matches**: `REF-*.md`, `*.pdf` added to research directories
3. **Agent activity**: Any agent writes to research corpus directories
4. **Manual trigger**: User requests source assessment
### Skip Conditions
- File is in `.aiwg/research/quality-assessments/` (already an assessment)
- File is `INDEX.md` or `README.md`
- File is a schema or template (`*.yaml` in schemas/)
- Assessment already exists for this REF-ID
## Behavior
When a new research source is detected:
1. **Extract metadata**
- Parse YAML frontmatter from source document
- Extract `ref_id`, `title`, `authors`, `year`, `source_type`
- If frontmatter missing, prompt agent to add it
2. **Determine baseline quality**
- Map source type to GRADE baseline:
- `peer_reviewed_journal` -> HIGH
- `peer_reviewed_conference` -> HIGH
- `preprint` -> MODERATE
- `technical_report` -> MODERATE
- `industry_whitepaper` -> LOW
- `blog_post` -> VERY LOW
- `forum_discussion` -> VERY LOW
3. **Invoke Quality Assessor**
- Delegate to Quality Assessor agent for full GRADE assessment
- Pass source metadata and content
- Request assessment in YAML format
4. **Store assessment**
- Save to `.aiwg/research/quality-assessments/{ref-id}-assessment.yaml`
- Update source frontmatter with `grade_level` field (if `--update-frontmatter`)
5. **Update corpus index**
- Add entry to quality assessment index
- Update GRADE distribution statistics
- Flag if corpus has > 30% unassessed sources
6. **Report**
- Display assessment summary to user
- Include hedging language recommendations
- Warn if source quality is LOW or VERY LOW
## Agent Orchestration
- **Primary**: Quality Assessor (performs the assessment)
- **Supporting**: Citation Verifier (validates existing citations of this source after assessment)
- **Notification**: Technical Writer, Documentation Synthesizer (if source is cited in existing docs, notify of GRADE level)
## Integration
### With Citation Guard
After assessment completes, Citation Guard uses the GRADE level to enforce hedging:
```yaml
integration:
citation_guard:
action: update_grade_cache
data: new_assessment
```
### With Research Metadata
Assessment populates fields required by research metadata rules:
```yaml
integration:
research_metadata:
fields_populated:
- quality_assessment.grade_level
- quality_assessment.baseline
- quality_assessment.downgrade_factors
- quality_assessment.upgrade_factors
```
### With Provenance Tracking
Assessment activity recorded in provenance chain:
```yaml
integration:
provenance:
activity_type: quality_assessment
agent: quality-assessor
```
## Configuration
```yaml
skill:
name: grade-on-ingest
type: passive
always_active_for:
- quality-assessor
- technical-researcher
- citation-verifier
file_triggers:
- pattern: ".aiwg/research/sources/REF-*.md"
- pattern: ".aiwg/research/findings/REF-*.md"
auto_assess: true
update_frontmatter: false # Requires --update-frontmatter flag
notify_on_low_quality: true
block_on_missing_frontmatter: false
```
## Output Locations
- Assessment: `.aiwg/research/quality-assessments/{ref-id}-assessment.yaml`
- Updated frontmatter: Source document (if `--update-frontmatter`)
- Index update: `.aiwg/research/quality-assessments/INDEX.md`
## References
- @$AIWG_ROOT/agentic/code/frameworks/sdlc-complete/agents/quality-assessor.md - Assessment agent
- @.aiwg/research/docs/grade-assessment-guide.md - GRADE methodology
- @$AIWG_ROOT/agentic/code/frameworks/sdlc-complete/schemas/research/quality-dimensions.yaml - Quality schema
- @$AIWG_ROOT/agentic/code/frameworks/sdlc-complete/rules/research-metadata.md - Metadata requirements
- @$AIWG_ROOT/agentic/code/frameworks/sdlc-complete/rules/citation-policy.md - Citation policy
- @$AIWG_ROOT/agentic/code/frameworks/sdlc-complete/skills/citation-guard/SKILL.md - Citation guardRelated Skills
memory-ingest
Ingest a source into any consumer's semantic memory by reading the topology contract
kb-ingest
Ingest a source (URL, file, or freeform note) into the knowledge base. Creates a source summary and updates or creates relevant entity and concept pages.
grade-report
Generate corpus-wide GRADE quality distribution report
aiwg-orchestrate
Route structured artifact work to AIWG workflows via MCP with zero parent context cost
venv-manager
Create, manage, and validate Python virtual environments. Use for project isolation and dependency management.
pytest-runner
Execute Python tests with pytest, supporting fixtures, markers, coverage, and parallel execution. Use for Python test automation.
vitest-runner
Execute JavaScript/TypeScript tests with Vitest, supporting coverage, watch mode, and parallel execution. Use for JS/TS test automation.
eslint-checker
Run ESLint for JavaScript/TypeScript code quality and style enforcement. Use for static analysis and auto-fixing.
repo-analyzer
Analyze GitHub repositories for structure, documentation, dependencies, and contribution patterns. Use for codebase understanding and health assessment.
pr-reviewer
Review GitHub pull requests for code quality, security, and best practices. Use for automated PR feedback and approval workflows.
YouTube Acquisition
yt-dlp patterns for acquiring content from YouTube and video platforms
Quality Filtering
Accept/reject logic and quality scoring heuristics for media content