القائمة
Track job applications, analyze positions, identify skill gaps, and plan your job search strategy
التثبيت باستخدام Codex أو Claude انسخ هذا Prompt والصقه في Codex أو Claude أو مساعد آخر ليراجع صفحة Skill ويثبّتها لك.
استنادا إلى تصنيف SOC المهني
Abductive argumentation-based literature trend analysis — trace how explanatory hypotheses evolve over time within a tagged literature thread
Personal health & fitness monitoring — ingest HealthKit metrics, track goals, surface regressions, manage appointments
Investigate rare disease mechanism of harm and therapeutic strategies from a known MONDO diagnosis
Credibility and consistency tracker for public figures. Ingests YouTube transcripts and news articles, extracts claims, detects contradictions, and builds position timelines.
| name | jobhunt |
| description | Track job applications, analyze positions, identify skill gaps, and plan your job search strategy |
| read_strategy | On first use: read Quick Start + Job Seeker Profile. When searching: read Discovery section (agent-driven search missions). When ingesting: read Ingestion + Sensemaking sections. When analyzing: read Analysis section. Full command reference: read Commands section on demand. |
| triggers | ["add job / ingest job / new position","search for jobs / run a search mission","analyze this job posting / make sense of","show my pipeline / skill gaps / learning plan","update status / job search","create seeker profile / set up job search"] |
| prerequisites | ["TypeDB running (make db-start)","make build-skills"] |
Use this skill to manage a job search as a knowledge graph. Claude acts as the career agent — browsing job boards, evaluating postings against the seeker's profile, and building structured understanding of positions, companies, and fit over time.
Key principle: The agent does the sensemaking. Scripts are tools that search APIs, store data, and run queries. The agent decides what's relevant, writes the rationale, and manages the pipeline.
Role model: The job seeker is a BFO/UFO role that inheres in a person. One person can have multiple roles (job seeker, author, evaluator). The role holds search-specific data; the person holds identity data.
make db-startuv sync --all-extras (from project root)make build-db after make build-skillsPath note: Replace
.claude/skills/jobhuntbelow with your installation directory:
- Claude plugin install:
${CLAUDE_PLUGIN_ROOT}/skills/jobhunt(self-contained bundle atplugins/jobhunt/)- skillful-alhazen project:
.claude/skills/jobhuntWhen installed as a plugin, TypeDB starts automatically on session start (no manual init required).
TYPEDB_HOST: TypeDB server (default: localhost)TYPEDB_PORT: TypeDB port (default: 1729)TYPEDB_DATABASE: Database name (default: alhazen_notebook)# Ingest a job posting from URL
uv run python .claude/skills/jobhunt/jobhunt.py ingest-job \
--url "https://boards.greenhouse.io/anthropic/jobs/123456" \
--priority high
# List your pipeline
uv run python .claude/skills/jobhunt/jobhunt.py list-pipeline
# Show position details
uv run python .claude/skills/jobhunt/jobhunt.py show-position --id "position-abc123"
# Add a note to any entity
uv run python .claude/skills/jobhunt/jobhunt.py add-note \
--about "position-abc123" --type research \
--content "Strong AI safety focus, good culture fit"
uv run emits a VIRTUAL_ENV warning to stderr. Always use 2>/dev/null when piping output to a JSON parser -- never 2>&1, which merges the warning into stdout and breaks JSON parsing.
The job seeker is modeled as a role (BFO/UFO pattern) that inheres in a person. The role holds search-specific data (target role, industries, salary expectations) while the person entity holds identity data (name, email, linkedin).
uv run python .claude/skills/jobhunt/jobhunt.py create-seeker-profile \
--person op-f25ab4b15b0f \
--target-role "Principal AI Architect" \
--industries "AI, Biotech, Scientific Computing" \
--salary "200k-250k" \
--location "Remote / SF Bay Area" \
--focus "TypeDB-powered knowledge systems, AI for science"
This creates a jhunt-job-seeker-role entity linked to the person via alh-role-bearing. All pipeline data (opportunities, skills, candidates) scopes to this role.
Skills are owned by the seeker role, derived from real evidence, and use a simplified Bloom's taxonomy for proficiency levels.
Proficiency levels:
| Level | Meaning | Evidence type | Interview readiness |
|---|---|---|---|
expert | Can design/teach | Publications, architected systems, led teams | Can whiteboard, critique alternatives |
practiced | Hands-on experience | Projects, production use, coursework | Can walk through how you used it |
aware | Conceptual knowledge | Read papers/docs, attended talks | Can answer "what is X" questions |
none | Not known | No evidence | Cannot discuss |
The agent ingests the seeker's professional artifacts as sources of evidence. Each artifact is stored and linked to the seeker via alh-aboutness so skills can trace back to their proof.
# Ingest LinkedIn profile (save page HTML or use Playwright MCP to capture)
uv run python .claude/skills/jobhunt/jobhunt.py ingest-job \
--url "https://linkedin.com/in/username"
# Ingest resume (PDF or text)
# Upload or paste resume content, agent stores as jhunt-resume artifact
# Ingest ORCID profile
# Agent fetches via WebFetch and stores key publications/experience
# Ingest Google Scholar page
# Agent fetches and extracts publication list, h-index, citation metrics
The agent should fetch each source, extract the content, and store it as an artifact linked to the seeker role via alh-aboutness. This creates the evidence chain: artifact → seeker → skill.
The agent reads the ingested artifacts and conducts a structured interview with the seeker:
For each skill identified, the agent assesses the proficiency level based on the evidence and discussion.
uv run python .claude/skills/jobhunt/jobhunt.py add-skill \
--name "Knowledge Graphs" --level "expert" \
--evidence "Built 5000+ entity TypeDB production KG, BFO/UFO ontology design" \
--recency "daily 2026" \
--description "TypeDB 3.x, TypeQL, schema evolution, GLAV mapping"
uv run python .claude/skills/jobhunt/jobhunt.py add-skill \
--name "RAG Pipelines" --level "expert" \
--evidence "Voyage AI + Qdrant in production, hybrid ontology-filtered retrieval" \
--recency "daily 2026"
uv run python .claude/skills/jobhunt/jobhunt.py add-skill \
--name "Pharma Experience" --level "none" \
--description "No direct pharma industry experience. Biomedical informatics at CZI."
Each skill is linked to the seeker via jhunt-seeker-has-skill. The --evidence field should reference the ingested artifacts where possible.
Once skills are populated, gap analysis joins seeker skills against position requirements:
uv run python .claude/skills/jobhunt/jobhunt.py show-gaps
uv run python .claude/skills/jobhunt/jobhunt.py list-skills
Discovery is agent-driven: Claude browses job boards, reads postings, evaluates fit against the seeker profile, and adds good matches as candidates with a written rationale. No automated scoring or LLM triage — the agent IS the sensemaker.
search-source) for bulk listing from Greenhouse/Lever boardssearch-source on website sources) for browsing JS-rendered job sites like hiring.cafeweb-search skill) for ad-hoc browsing of careers pages or niche boardsuv run python .claude/skills/jobhunt/job_forager.py add-candidate \
--title "Research Scientist, AI for Science" \
--url "https://boards.greenhouse.io/anthropic/jobs/123" \
--location "San Francisco" \
--relevance 0.85 \
--reason "Strong fit: combines AI evaluation with scientific methodology"
Three types of sources:
# Company recruiting pages (Greenhouse/Lever/Ashby) — API-based, returns all jobs
uv run python .claude/skills/jobhunt/job_forager.py add-source \
--name "Anthropic" --platform greenhouse --token anthropic
# Job board aggregators (LinkedIn/Remotive/Adzuna) — API-based, keyword search
uv run python .claude/skills/jobhunt/job_forager.py add-source \
--name "ML Jobs" --platform linkedin --query "machine learning" --location "San Francisco"
# Job websites (hiring.cafe, etc.) — browser-based via Playwright MCP
uv run python .claude/skills/jobhunt/job_forager.py add-source \
--name "hiring.cafe" --platform website \
--url "https://hiring.cafe" \
--query "AI scientist knowledge graph biotech drug discovery"
| Platform | Type | Auth | Args |
|---|---|---|---|
greenhouse | Company board | None | --token (slug) |
lever | Company board | None | --token (slug) |
ashby | Company board | None | --token (slug) |
linkedin | Aggregator | None | --query, --location |
remotive | Aggregator | None | --query, --location |
adzuna | Aggregator | API key | --query, --location |
website | Job website | Playwright MCP | --url, --query |
Returns all job listings from a source for the agent to evaluate:
uv run python .claude/skills/jobhunt/job_forager.py search-source --source "Anthropic"
# Returns: {jobs: [{title, url, location, external_id}, ...]}
The agent reads through these, fetches interesting postings via web-fetch, and adds good matches as candidates.
For website platform sources, search-source returns instructions instead of job listings.
The agent uses Playwright MCP tools to browse the site interactively:
search-source --source "hiring.cafe" returns the URL and querybrowser_navigate to the source URLbrowser_type the query into the search box, submitbrowser_snapshot or browser_take_screenshot to see rendered resultsbrowser_run_code_unsafe to extract job URLs from the DOM, or read the snapshotWebFetch each interesting job URL (individual pages usually render server-side)add-candidate with rationale for each good matchExample: hiring.cafe
hiring.cafe is a 41M+ job aggregator with AI-powered search. It uses client-side React rendering, so API/curl cannot access search results — Playwright MCP is required.
The site auto-generates filters from natural language queries (Department, Industry, Tech Keywords). Sometimes the auto-filters are too narrow and return zero results. If this happens, click "Reject All" on the suggested filters and try a broader query, or remove specific filter chips.
Individual job pages (/viewjob/{id}) render server-side and can be fetched with WebFetch for full details.
Candidates are embedded in Qdrant on creation. Search by meaning:
uv run python .claude/skills/jobhunt/job_forager.py search-candidates \
--query "knowledge graph engineering for science" --limit 10
# List all candidates
uv run python .claude/skills/jobhunt/job_forager.py list-candidates
# List by status
uv run python .claude/skills/jobhunt/job_forager.py list-candidates --status reviewed
# Promote a candidate to a full position (creates jhunt-position + application note)
uv run python .claude/skills/jobhunt/job_forager.py promote --id candidate-abc123
The agent does the sensemaking. Scripts are tools — they search APIs, store data, and run queries. The agent decides what's relevant, writes the rationale, and manages the pipeline.
Triggers: "add job", "ingest job", "new position", "found a job posting", "here's a job"
uv run python .claude/skills/jobhunt/jobhunt.py ingest-job \
--url "https://boards.greenhouse.io/anthropic/jobs/123456" \
--priority high \
--tags "ai" "ml" "safety"
Options:
--url (required): Job posting URL--priority: Set priority (high/medium/low)--tags: Space-separated tagsReturns:
{
"success": true,
"position_id": "position-abc123",
"artifact_id": "artifact-xyz789",
"status": "raw",
"message": "Artifact stored - ask Claude to 'analyze this job posting' for sensemaking."
}
What ingestion produces:
jobhunt-position entity with status researchingjobhunt-job-description artifact containing the raw scraped page contentjobhunt-application-note tracking the application statusuv run python .claude/skills/jobhunt/jobhunt.py add-position \
--title "Senior ML Engineer" \
--company "Anthropic" \
--priority high
uv run python .claude/skills/jobhunt/jobhunt.py add-company \
--name "Anthropic" \
--url "https://anthropic.com" \
--description "AI safety research company"
uv run python .claude/skills/jobhunt/jobhunt.py add-engagement \
--name "Acme Corp Data Consulting" \
--company-id "company-abc123" \
--type project \
--rate "$200/hr" \
--status active \
--priority high
Engagement types: hourly | project | retainer | advisory
uv run python .claude/skills/jobhunt/jobhunt.py add-venture \
--name "Augura Health Advisory" \
--stage proposal-sent \
--equity-type advisor \
--priority high
Venture stages: exploring | proposal-sent | negotiating | active | paused | closed
Equity types: none | advisor | cofounder | investor
uv run python .claude/skills/jobhunt/jobhunt.py add-lead \
--name "Jane Smith - BigCo" \
--status warm \
--priority medium \
--description "Met at ML Summit, interested in consulting"
Lead statuses: first-contact | active | inactive | closed
When user says "analyze this job posting" or "make sense of [position]":
Get the artifact content
uv run python .claude/skills/jobhunt/jobhunt.py show-artifact --id "artifact-xyz"
Read and comprehend the content
Research the company and leadership online (use the web-search skill)
Search for: company mission, funding, recent news, and LinkedIn activity of key leaders. Focus on:
Save findings as a research note attached to both the company and the position:
# Company-level research (background, funding, mission)
uv run python .claude/skills/jobhunt/jobhunt.py add-note \
--about "company-xyz" \
--type research \
--content "Series C, $4B raised. Mission: X. CEO recently spoke at Y conference..."
# Position-level research (leadership context, role fit signals, hiring manager)
uv run python .claude/skills/jobhunt/jobhunt.py add-note \
--about "position-abc123" \
--type research \
--content "Bo Wang (SVP Biomedical AI) is likely hiring manager. Very active on LinkedIn..."
Create/update the company
uv run python .claude/skills/jobhunt/jobhunt.py add-company \
--name "Anthropic" \
--url "https://anthropic.com" \
--description "AI safety research company"
Extract requirements using the skill vocabulary
Before extracting requirements, load the skill concept vocabulary:
uv run python .claude/skills/jobhunt/jobhunt.py list-concepts
This returns the full vocabulary with your proficiency levels, alt-labels, and hierarchy. Use canonical concept names when creating requirements so they match your skill profile.
For each skill in the posting:
# Create a new concept if needed
uv run python .claude/skills/jobhunt/jobhunt.py add-concept \
--name "Molecular Simulation" \
--alt-labels "MD,Molecular Dynamics,Force Fields"
# Add the user's self-assessed skill
uv run python .claude/skills/jobhunt/jobhunt.py add-skill \
--name "Molecular Simulation" --level "aware" \
--evidence "Read papers, understand concepts, no hands-on" \
--recency "reading 2026"
# Then add the requirement (using the canonical concept name)
uv run python .claude/skills/jobhunt/jobhunt.py add-requirement \
--position "position-abc123" \
--skill "Molecular Simulation" \
--level "required"
Key principle: Never assume the user's skill level. If a requirement mentions a skill not in their profile, ask them. Their self-assessment + evidence is the ground truth, not the agent's inference.
Create analysis notes
Fit Analysis Note:
uv run python .claude/skills/jobhunt/jobhunt.py add-note \
--about "position-abc123" \
--type fit-analysis \
--content "Strong fit for core requirements. Gap in distributed systems." \
--fit-score 0.82 \
--fit-summary "Strong technical fit, one gap to address"
Skill Gap Note:
uv run python .claude/skills/jobhunt/jobhunt.py add-note \
--about "position-abc123" \
--type skill-gap \
--content "Distributed systems is required. Recommend: DDIA book, MIT 6.824 course."
Flag uncertainties
uv run python .claude/skills/jobhunt/jobhunt.py tag \
--entity "requirement-xyz" \
--tag "uncertain"
Report findings to user: company overview, leadership signals, fit score breakdown, key gaps, suggested next steps (including who to follow on LinkedIn)
uv run python .claude/skills/jobhunt/jobhunt.py list-artifacts --status raw
uv run python .claude/skills/jobhunt/jobhunt.py list-artifacts --status all
uv run python .claude/skills/jobhunt/jobhunt.py show-artifact --id "artifact-xyz789"
Every opportunity MUST have after sensemaking:
--company flag on add-position, auto-matched to existing jobhunt-company entities (do not create duplicates)salary-range attributejobhunt-requirement entities via add-requirementfit-score (0.0-1.0) and fit-summaryjobhunt-opp-summary-note synthesizing all notes. Regenerated after any note is added or updated.embedding_map.py embed-and-map after saving the summary so the opportunity appears on Mission Control.Every opportunity has exactly one jobhunt-opp-summary-note — a living markdown dossier that is overwritten each time notes change. This is the primary embedding text for the Mission Control map and the quick-read view for understanding any opportunity.
Workflow — regenerate after any note update:
# 1. Fetch all notes + metadata
uv run python .claude/skills/jobhunt/jobhunt.py regenerate-summary --about <opp-id>
# 2. Read the JSON output, write a markdown summary following the template below
# 3. Save the summary (creates or overwrites)
uv run python .claude/skills/jobhunt/jobhunt.py upsert-summary --about <opp-id> --content "@/tmp/summary.md"
# 4. Re-embed to update the Mission Control map
uv run python local_skills/jobhunt/embedding_map.py embed-and-map
IMPORTANT: Step 4 (re-embed) MUST be run after saving the summary. Without it, new or updated opportunities will not appear on the Mission Control map. This step re-computes embeddings for ALL opportunities and regenerates the 2D layout.
Summary templates by type:
Position:
## Role
- Title, company, location/remote policy
- Key responsibilities (2-3 bullets)
- Salary range if known
## Fit
- Overall fit score and one-line assessment
- Top strengths (2-3 bullets with specifics)
- Key gaps (1-2 bullets)
## Company
- What they do, stage/size, why interesting
- Key people (hiring manager, contacts)
## Status
- Current application status
- Key dates, next steps, or outcome
Engagement:
## Engagement
- Client, scope, type (consulting/contract/advisory)
- Rate/compensation if known
## Fit
- Why this is a good match
- Key deliverables
## Status
- Current stage (proposal/active/paused/closed)
Venture:
## Overview
- What the venture is, stage
- Your role/involvement
## Opportunity
- Why it's interesting
- Key milestones or next steps
## Status
- Current stage (seed/series-a/series-b/growth/closed)
Lead:
## Contact
- Who, title, organization
- How you met, when
## Context
- What the connection is about
- Potential opportunity or value
## Status
- Relationship state (first-contact/active/inactive/closed)
uv run python .claude/skills/jobhunt/jobhunt.py update-status \
--position "position-abc123" \
--status "applied" \
--date "2025-02-05"
Status values: researching | applied | phone-screen | interviewing | offer | rejected | withdrawn
# Interaction note
uv run python .claude/skills/jobhunt/jobhunt.py add-note \
--about "position-abc123" \
--type interaction \
--content "Phone screen went well, moving to technical round." \
--interaction-type "call" \
--interaction-date "2025-02-05"
# Strategy note
uv run python .claude/skills/jobhunt/jobhunt.py add-note \
--about "position-abc123" \
--type strategy \
--content "Lead with distributed systems experience from caching project."
# Claude Code brief (CC agent writes a brief for the next session)
uv run python .claude/skills/jobhunt/jobhunt.py add-note \
--about "position-abc123" \
--type cc-brief \
--content "Next session: prep for technical interview on 2025-02-10. Focus on system design questions."
# Claude Code feedback (CC agent records feedback from completed interaction)
uv run python .claude/skills/jobhunt/jobhunt.py add-note \
--about "position-abc123" \
--type cc-feedback \
--content "User reported phone screen went well. Interviewer asked about distributed caching. Move to next round."
Note types: research | strategy | interview | interaction | skill-gap | fit-analysis | general | cc-brief | cc-feedback
uv run python .claude/skills/jobhunt/jobhunt.py show-gaps
uv run python .claude/skills/jobhunt/jobhunt.py learning-plan
uv run python .claude/skills/jobhunt/jobhunt.py add-resource \
--name "Designing Data-Intensive Applications" \
--type "book" \
--url "https://dataintensive.net" \
--hours 30 \
--skills "distributed-systems" "system-design"
uv run python .claude/skills/jobhunt/jobhunt.py link-resource \
--resource "<resource-id>" \
--requirement "<requirement-id>"
# Search for papers on a skill gap topic
uv run python .claude/skills/scientific-literature/scientific_literature.py search \
--query "machine learning systems design" \
--collection "ML Systems Reading List"
# Link collection to skill gap
uv run python .claude/skills/jobhunt/jobhunt.py link-collection \
--collection "<collection-id>" \
--skill "machine-learning"
# Link a specific paper to a learning resource
uv run python .claude/skills/jobhunt/jobhunt.py link-paper \
--resource "<resource-id>" \
--paper "<paper-id>"
# View updated plan
uv run python .claude/skills/jobhunt/jobhunt.py learning-plan
# All positions
uv run python .claude/skills/jobhunt/jobhunt.py list-pipeline
# Filter by status or priority
uv run python .claude/skills/jobhunt/jobhunt.py list-pipeline --status "interviewing"
uv run python .claude/skills/jobhunt/jobhunt.py list-pipeline --priority "high"
# All opportunity types
uv run python .claude/skills/jobhunt/jobhunt.py list-opportunities --type all
uv run python .claude/skills/jobhunt/jobhunt.py list-opportunities --type venture
uv run python .claude/skills/jobhunt/jobhunt.py list-opportunities --type engagement --status active
# Detail views
uv run python .claude/skills/jobhunt/jobhunt.py show-position --id "position-abc123"
uv run python .claude/skills/jobhunt/jobhunt.py show-opportunity --id "venture-abc123"
uv run python .claude/skills/jobhunt/jobhunt.py show-company --id "company-xyz"
uv run python .claude/skills/jobhunt/jobhunt.py show-gaps
uv run python .claude/skills/jobhunt/jobhunt.py learning-plan
uv run python .claude/skills/jobhunt/jobhunt.py report-pipeline # Pipeline overview
uv run python .claude/skills/jobhunt/jobhunt.py report-stats # Stats summary
uv run python .claude/skills/jobhunt/jobhunt.py report-gaps # Skill gaps
uv run python .claude/skills/jobhunt/jobhunt.py report-position --id "position-xyz" # Position detail
Use reports (Markdown) for displaying to users in chat. Use JSON commands (list-pipeline, show-position) for programmatic processing.
uv run python .claude/skills/jobhunt/jobhunt.py tag --entity "position-abc123" --tag "remote"
uv run python .claude/skills/jobhunt/jobhunt.py search-tag --tag "remote"
The jobhunt skill tracks multiple types of career opportunities via the jobhunt-opportunity hierarchy:
jobhunt-opportunity (base)
+-- jobhunt-position -- formal employment application (has application-status pipeline)
+-- jobhunt-engagement -- paid consulting/service work for a client
+-- jobhunt-venture -- startup/advisory/equity opportunity
+-- jobhunt-lead -- early-stage networking contact, role undefined
All opportunity types share: opportunity-status, priority-level, deadline, and can be linked to a jobhunt-company via opportunity-at-organization.
All opportunity types work with add-note --about <ID> -- notes attach to any identifiable-entity.
uv run python .claude/skills/jobhunt/jobhunt.py update-opportunity \
--id "venture-abc123" \
--status active \
--stage negotiating \
--priority high
| Situation | Entity type | Key attributes |
|---|---|---|
| Own consulting business | jobhunt-company | name, description |
| Startup advisory role | jobhunt-venture | venture-stage, equity-type |
| Consulting engagement | jobhunt-engagement | engagement-type, rate-info |
| Networking contact (no role yet) | jobhunt-lead | opportunity-status, description |
| Formal job application | jobhunt-position | job-url, application-status pipeline |
Phylo strategy pattern: Use an existing jobhunt-position as the anchor for all interactions. Add jobhunt-interaction-note entries for each meeting. Only create a separate jobhunt-lead if a genuinely new thread emerges from those conversations.
| Type | Description |
|---|---|
your-skill | Your skills for gap analysis |
jobhunt-company | An employer/client organization |
jobhunt-opportunity | Base type for all opportunities |
jobhunt-position | Formal job posting (sub opportunity) |
jobhunt-engagement | Consulting engagement (sub opportunity) |
jobhunt-venture | Startup/advisory venture (sub opportunity) |
jobhunt-lead | Networking lead (sub opportunity) |
jobhunt-learning-resource | Course, book, tutorial |
jobhunt-contact | Person at a company |
jobhunt-search-source | Company board for forager |
jobhunt-candidate | Discovered posting (forager) |
| Type | Description |
|---|---|
jobhunt-job-description | Raw HTML/text from job posting URL |
jobhunt-resume | Resume document |
jobhunt-cover-letter | Cover letter |
jobhunt-company-page | Company website snapshot |
jobhunt-proposal | Proposal or pitch deck for engagement/venture |
| Type | Purpose |
|---|---|
jobhunt-application-note | Status tracking (positions) |
jobhunt-research-note | Company/opportunity research |
jobhunt-interview-note | Interview prep/feedback |
jobhunt-strategy-note | Talking points, approach |
jobhunt-skill-gap-note | Learning needs |
jobhunt-fit-analysis-note | Fit assessment |
jobhunt-interaction-note | Contact logs |
jobhunt-cc-brief-note | Claude Code brief for next session |
jobhunt-cc-feedback-note | Claude Code feedback from completed interaction |
position-at-company -- links position to employeropportunity-at-organization -- links any opportunity type to a companyaboutness -- links notes to any entity (position, company, opportunity)requirement-of -- links requirements to positionslocal_skills/jobhunt/schema.tqllocal_resources/typedb/alhazen_notebook.tql| Command | Description | Key Args |
|---|---|---|
ingest-job | Fetch job URL, store raw | --url |
add-skill | Add/update skill with Bloom's proficiency | --name, --level (expert/practiced/aware/none), --evidence, --recency |
list-skills | Show your skills | |
list-artifacts | List artifacts by status | --status |
show-artifact | Get artifact content | --id |
add-company | Add company | --name |
add-position | Add position manually | --title |
add-engagement | Add consulting engagement | --name, --type, --rate |
add-venture | Add startup/advisory venture | --name, --stage, --equity-type |
add-lead | Add networking lead | --name, --status |
update-opportunity | Update any opportunity status/stage/priority | --id |
show-opportunity | Show any opportunity details | --id |
list-opportunities | List by type/status | --type, --status, --priority |
add-requirement | Add skill requirement | --position, --skill |
update-status | Change position application status | --position, --status |
add-note | Create any note type | --about, --type, --content |
add-resource | Add learning resource | --name, --type |
link-collection | Link paper collection to skill gap | --collection, --skill |
link-resource | Link resource to a skill requirement | --resource, --requirement |
link-paper | Link learning resource to a paper | --resource, --paper |
create-seeker-profile | Create job-seeker role for a person | --person, --target-role, --industries |
list-pipeline | Show position pipeline | --status, --priority, --person |
show-position | Position details | --id |
show-company | Company details | --id |
show-gaps | Skill gap analysis | |
learning-plan | Prioritized study list | |
tag | Tag an entity | --entity, --tag |
search-tag | Find by tag | --tag |
report-pipeline | Pipeline overview (Markdown) | |
report-stats | Stats summary (Markdown) | |
report-gaps | Skill gaps report (Markdown) | |
report-position | Position detail (Markdown) | --id |
| Command | Description | Key Args |
|---|---|---|
add-source | Add a search source (company board or aggregator) | --name, --platform, --token/--query |
list-sources | List configured search sources | |
remove-source | Remove a source | --id, --token, or --name |
search-source | Search one source (raw listings for agent review) | --source |
add-candidate | Add a candidate with agent evaluation | --title, --url, --relevance, --reason |
search-candidates | Semantic search across candidates (Qdrant) | --query, --limit |
list-candidates | List candidates | --status, --source |
promote | Promote candidate to full position | --id |
Declarative audit rules for data quality are defined in quality-checks.yaml.
# Check for violations
uv run python .claude/skills/jobhunt/jobhunt.py audit
# Auto-fix where possible
uv run python .claude/skills/jobhunt/jobhunt.py audit --fix
# Or use the generic audit runner
uv run python src/skillful_alhazen/utils/audit_runner.py run \
--checks local_skills/jobhunt/quality-checks.yaml
| Check | Severity | Description |
|---|---|---|
position-company-link | high | Positions should be linked to a company via position-at-company |
ugly-titles | medium | Position titles containing job-board boilerplate |
missing-short-name | medium | Positions without a short-name for compact display |
missing-salary | low | Positions without salary-range information |
positions-without-notes | low | Positions with no notes at all |
duplicate-companies | high | Multiple company entities with identical names |
orphaned-companies | medium | Companies with no positions or opportunities linked |
A Next.js dashboard is available for visualizing your job search pipeline. It runs in Docker (not npm run dev).
docker compose build --no-cache dashboard
docker compose up -d dashboard
# Dashboard available at http://localhost:3001
/jobhunt) -- Kanban columns by application status/jobhunt/position/{id}) -- Full position profile: requirements, notes, gap analysis, fit score/jobhunt/collection/{id}) -- Notes and resources grouped by collectiondashboard/src/app/(jobhunt)/jobhunt/dashboard/src/components/jobhunt/dashboard/src/app/api/jobhunt/dashboard/src/lib/jobhunt.tsDuring sensemaking, if you encounter a concept, relationship, or entity type that has no place in the current TypeDB schema, that is a schema gap -- a signal for schema evolution, not a failure.
When you notice a schema gap:
uv run python local_resources/skilllog/skill_logger.py file-schema-gap \
--skill jobhunt \
--concept "<the concept you tried to represent>" \
--missing "<which TypeDB entity/relation/attribute is absent>" \
--suggested "<proposed TypeQL addition, or 'unknown' if unsure>"
Examples of schema gaps:
jobhunt-contactUse --dry-run first to review the issue before filing it.