67-subagent extraction engine
Career Talent Extractor
Turn your professional artifacts — code, resumes, git history, Slack, transcripts — into a structured career knowledge base and ready-to-use job-search content.
Phase‑0 provenance separates what you did from what you facilitated
Built to find the evidence you would miss
CTE is an extraction engine, not a resume template. It reads your real work and reconstructs the career narrative hiding inside it.
How it works
See the full architectureDependency-ordered tiers, 67 prompts
Each tier consumes the tier below it. Facts come first; the meta tier reasons across all of them last.
Execution tiers
- T1FactualEstablishes ground-truth facts
- T2Hard SkillsTechnical depth & craft
- T3Soft SkillsBehavioral evidence
- T4NarrativeStories & impact arcs
- T5Role LensPer-role fit signals
- T6IndustryDomain expertise
- T7MetaCross-references everything
- T0Content GenPost-aggregation deliverables
Subagent categories
- Quantitative6
- Temporal4
- Skill Domain9
- Soft Skills9
- Narrative9
- Role-Based11
- Industry7
- Meta6
- Content Generation6
Four commands, start to offer
CTE runs inside Claude Code or Cowork. Onboard once with /init, then extract, synthesize, and generate — no setup beyond the plugin.
- 1Capture your profile, target roles, and sources
/init - 2Run the engine over an artifact
/extract ./path - 3Build your talent profile
/synthesize - 4Generate ready-to-use content
/advise