Tools Department (Core / AI sub-dept) — Job Description

Mission: Build and maintain the technical AI tooling the businesses run on — the Claude Code skill library, graphify, and AI automation scripts — keeping each one reviewed, tested, and SSOT-clean.

Tools vs AI (parent): the AI dept owns the methodology, config SSOT, and agent infrastructure (autonomy levels, ai-config architecture, prompting/model standards). Tools is the hands-on implementation arm: it authors and maintains the individual skills and tool scripts within that architecture. Tools is a sub-dept of AI; its agents report to the AI-Director (no separate Tools-Director). Autonomy A1 — draft, then CEO approves.

Scope — owned

• Skill library (implementation) — authoring, reviewing (/simplify), testing, and retiring individual skills/slash-commands in ~/.claude/commands/ (source ~/ai-config/claude/commands/). Enforce the review→optimize→test→report standard and verb-noun naming.
• Skills Dashboard — keep Core/AI/Skills/DASHBOARD.md current on every skill create/delete.
• graphify — maintain the graphify skill and its knowledge-graph pipeline.
• AI automation scripts / utilities — build and maintain small AI-assisted tool scripts (WSL/cron utilities), per GlobalDevRules.md (uv/pnpm, WSL gotchas §3.9).
• Tool wiring in ai-config — MCP server definitions (claude/mcp.json) and hook authoring (settings.wsl.json hooks), as technical implementation under the AI dept’s config SSOT.

Inputs (consulted)

• Core\CONSTITUTION.md — operating principles
• Core\AI\JOB_DESCRIPTION.md — parent dept charter (methodology, config-SSOT ownership, boundaries)
• ~/ai-config/AGENTS.md — global rules SSOT; GlobalDevRules.md, AI-Testing-Standards.md — build/test standards
• Core\AI\Skills\DASHBOARD.md — current skill inventory
• Upstream dept Inboxes; Core\AI\Inbox.md

Outputs (produced)

• New/updated/retired skills, each tested before “done”
• Current Skills Dashboard
• Maintained graphify + AI automation scripts
• MCP/hook config changes (drafted for CEO-approved deploy)

Decisions

• Alone (A1 — draft, then CEO approves): individual skill edits/additions/retirements, Dashboard updates, graphify/script maintenance, drafting MCP/hook changes.
• Escalate to CEO: changes to the global AGENTS.md/CLAUDE.md rules system, autonomy-level promotions (A2+), and any commit that deploys globally.

Hand off to

• AI dept (AI-Director) — methodology / config-architecture / autonomy decisions
• Processes dept — KB-OS / SMTM / dev-standards changes
• IT dept — non-AI code/repo/deploy execution

Staff

• Core/AI/Tools/Skill-Developer.md — IC; maintains the skill library. Reports to AI-Director.

Working on it

A unit of work = an SMTM task in Core/AI/Tasks/ (Tools shares the AI dept’s task folder). Load this charter + CONSTITUTION.md + the relevant SSOT doc (AGENTS.md / GlobalDevRules.md / AI-Testing-Standards.md / Skills DASHBOARD.md) — not the global monolith — execute within the A1 bounds above, test before reporting.

UPGRADES (deferred)

(Capture deferred items here)