vitaecontext-vitaegraph
Build, deepen, validate, index, and maintain a private hierarchical career knowledge graph from supplied career materials. Use when the user asks to create or update a VitaeGraph, model education with nested courses or thesis work, enrich projects from Git repositories, or supply deep selected career context to another VitaeContext skill.
vitaecontext · Static checks on file
Installation
Read the source and dependencies before running a command. Copying does not install or execute anything.
npx skills add vitaecontext/vitaecontext --skill vitaecontext-vitaegraphAbout this skill
name: vitaecontext-vitaegraph description: Build, deepen, validate, index, and maintain a private hierarchical career knowledge graph from supplied career materials. Use when the user asks to create or update a VitaeGraph, model education with nested courses or thesis work, enrich projects from Git repositories, or supply deep selected career context to another VitaeContext skill. license: MIT metadata: homepage: https://vitaecontext.github.io/ repository: https://github.com/vitaecontext/vitaecontext
VitaeContext VitaeGraph
Overview
Act as a career knowledge architect and diligent researcher. Turn supplied material into a deep, navigable graph that adds substantial detail beyond a compact career context file. Favor coherent domain modeling and useful prose over shallow coverage.
Progressive workflow selection
Read only the node workflows supported by the source inventory:
- Education, degrees, university courses, or thesis: node-workflows/education.md
- Projects or Git repository URLs: node-workflows/projects.md
- Employment, internships, volunteering, or research roles: node-workflows/experience.md
- Certifications or non-degree training: node-workflows/certifications.md
- Awards or recognition: node-workflows/awards.md
- Papers, talks, or publications: node-workflows/publications.md
Read references/record-workflow.md for paths, IDs, hierarchy, and graph maintenance. Read references/retrieval-and-handoff.md only when selecting existing graph context for downstream work. Read references/maintenance-and-migration.md when correcting, moving, merging, splitting, or deleting existing records.
Wiki context
- Read wiki/index.md for graph structure, validation, indexing, migration, privacy, or retrieval behavior.
- Read wiki/knowledge.md only after the index routes the task there.
- If wiki files are unavailable, continue with the matching reference and avoid stronger architecture claims.
Task mode
Select exactly one primary mode before loading node workflows:
Create: initialize an absent graph and build records from explicitly supplied material.Deepen: add supported detail to selected existing records without restructuring unrelated domains.Maintain: correct facts, repair relationships, merge or split records, move paths, or remove stale material.Validate: inspect graph structure and run validation without rewriting substantive content unless the user asks for repair.Index: validate first, then rebuild generated retrieval artifacts without changing canonical Markdown.Retrieve: select the smallest relevant record set for a downstream task without modifying the graph.Migrate: preview and then apply a deliberate hierarchy or path change while preserving stable IDs.
For Retrieve, read only references/retrieval-and-handoff.md plus wiki context when needed. For Maintain or Migrate, read references/maintenance-and-migration.md. Load node workflows only for record types whose content will be created or materially deepened.
Depth contract
Use the smallest graph pass that satisfies the selected mode:
Record pass: one named record and the relationships required to keep it valid.Domain pass: one supported domain, its children, cross-links, root/index summaries affected by it, and validation.Graph pass: all supplied domains, root synthesis, index, validation, and indexing.
Default to a record pass for narrow maintenance and a domain pass for creation from one source group. Use a graph pass only when the user asks to build, reconcile, migrate, or validate the whole graph.
Intake and authority
- Treat an explicit request to create, update, deepen, repair, migrate, or index a named graph as authority for the corresponding scoped mutations.
- Treat audit, explain, retrieve, and validate requests as read-only unless structural repair is also requested.
- Resolve and report the exact graph path before the first write. Use
~/.vitaecontext/vitaegraphonly when the user did not supply another path. - Inspect only explicit sources. Do not search for other career files or graphs.
- Never use
--force, replace root templates in a non-empty graph, delete a record, or perform a many-record migration without previewing the affected paths and obtaining explicit approval.
Create and deepen workflow
- Resolve the graph path. Use
~/.vitaecontext/vitaegraphunless the user supplied an exact directory. - Inspect all explicitly supplied sources before creating records. Do not scan unrelated filesystem locations.
- Produce an internal graph blueprint: available domains, proposed nodes, parent-child placement, cross-links, enrichment actions, and material gaps.
- Initialize the graph when absent. Never replace a non-empty graph or use
--forcewithout approval. - Process one domain at a time. Finish its applicable node workflow, cross-links, and completeness pass before switching domains.
- For every node, loop through extraction, enrichment, synthesis, relationship linking, and gap review. A title plus a short summary is not a finished node.
- Update
index.mdafter detailed records exist, then synthesizeVITAEGRAPH.mdfrom the completed graph. - Run graph validation. Repair structural errors within the authorized scope. Run graph indexing only after validation passes.
Do not spend user-facing tokens narrating the blueprint unless the user asks. Use it to structure execution.
Graph rules
- Store canonical user data in Markdown and generated JSON only under
.generated/. - Keep `t