ai-company

Micro-tier governance skill for small AI agent organizations (1-3 agents). Consolidates 18 enterprise function blocks into two departments, governance-and-delivery and engineering-and-safety, covering strategy, risk, compliance, architecture, safety, and delivery. Ships ten harness code templates, three prompt frameworks, a full error-code catalog, and a scaling specification. Provides two self-contained human-paste prompts (method implementation, robustness checks) usable in any AI chat window. Source files are authored in English; runtime output follows the user's language per the language policy, with English fallback. Includes propose-only self-scaling that evaluates scaling thresholds and proposes tier upgrades up to group. Use when a small team needs enterprise-grade governance, engineering harnesses, PII masking, AIGC disclosure, or tier-upgrade assessment for its agents.

AI Company Governance

Micro-tier (XS) LLM agent governance skill for organizations of 1-3 agents: 18 enterprise function blocks consolidated into 2 departments, backed by shared harness templates, a full error-code catalog, and propose-only self-scaling.

Source files are authored in English; runtime output follows the user's language (see language_policy in the frontmatter). Slugs, error codes, field names, enum values, file paths, and function names are never translated.

Departments

Slug	Prefix	Scope	Reference
governance-and-delivery	CEO_	Strategy, governance, delivery	governance-and-delivery.md
engineering-and-safety	CTO_	Architecture, engineering, safety	engineering-and-safety.md

Shared Resources

Resource	Contents
method-patterns.md	Single source of truth for the ten shared code templates and the CRISPE / 3WEH / Five-Element prompt frameworks
error-codes.md	Full four-element error-code catalog (code, message, trigger, resolution) for all prefixes
scaling.md	Tier thresholds, upgrade gating, and alias mapping for self-scaling

Error Code Prefixes

Prefix	Meaning
CEO_	Governance-and-delivery department codes
CTO_	Engineering-and-safety department codes
SCL_	Self-scaling infrastructure codes (fixed across tiers)

Prompts

File	Mode	Purpose
01-implement-method.md	human-paste	Paste into any AI chat window to have that AI implement one method to this standard
02-robustness-checks.md	human-paste	Paste into any AI chat window to have that AI run robustness checks on a method

Quick Start

1. Invoke the skill with task (required) and optionally department (auto routes automatically); set output_language to override the runtime output language.
2. For implementation standards, use the shared templates in method-patterns.md; for failures, look up codes by prefix in error-codes.md.
3. To work outside this skill, open 01-implement-method.md or 02-robustness-checks.md, fill in the <placeholders>, and paste the prompt into any AI chat window.
4. When the organization grows, ask the skill to evaluate scaling thresholds or propose a tier upgrade; the rules live in scaling.md.