citadel-mason
Use when a user wants to create, customize, export, import, or review a GoatCitadel Citadel Blueprint.
Skill by goatcitadel
Install
Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.
npx agentmarket install skill/goatcitadel-citadel-mason
Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.
Get the weekly diff
The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.
About
Reviews
Write a reviewNo reviews yet. Installed it? Reviews come from verified installers only. How reviews work
Permissions
What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.
- Tools
- No tools found
- Secrets
- None asked for
- Install scripts
- None found
Network hosts it talks to 1
codeload.github.com· found in install command
Files it touches
~/.claude/skills/citadel-mason
What's inside
- Skill name
citadel-mason- Description
- Use when a user wants to create, customize, export, import, or review a GoatCitadel Citadel Blueprint. Produces safe, schema-valid `goatcitadel.blueprint.v1` Blueprints for protected AI operating spaces.
- License
- Apache-2.0
- Instructions
- 77 lines in SKILL.md · view on GitHub
Files (1)
- SKILL.md
SKILL.md (first lines)
# Citadel Mason Skill
The Mason helps design **Citadels** — protected AI operating spaces — by producing
safe, reviewable **Blueprints**. A Blueprint is portable structure only: a Charter
and Chambers. It never contains secrets, credentials, tokens, private data, or
live authorization.
See `docs/citadel-blueprint-schema.md` for the full `goatcitadel.blueprint.v1` schema.
## When to use
The user wants to create, customize, export, import, or review a Citadel Blueprint —
or set up a new Citadel for a part of their life or work.
## Core rules (non-negotiable)
1. **Never** include secrets, API keys, OAuth/refresh tokens, private files, or credentials.
2. **Never** enable external write actions by default. Propose permissions; do not assume them.
3. Use **Chambers** for sensitive areas; mark them `sealed: true` with a `restricted`/`sensitive`/`secret` sensitivity.
4. Use conservative Gatehouse defaults (`riskPosture: "conservative"` for company/finance/legal contexts).
5. Output a **schema-valid** `goatcitadel.blueprint.v1` (see schema doc) — it must pass `validateCitadelBlueprint`.
6. Always include a human-readable **review summary** and `riskNotes`.
7. A Blueprint is **staged configuration, not live activation.** Only the human opens Gates.
## Workflow
1. Ask what the Citadel is for, and pick the `kind` (`personal`/`company`/`project`/…).
2. Ask what is sensitive → those become `sealed` Chambers.
3. Ask what success looks like → `charter.successDefinition`.
4. Ask what must always require approval / never happen → `charter.boundaries`.
5. Draft the Charter (`purpose`, `goals`, `boundaries`, `successDefinition`, `riskPosture`, `modelPolicyDefault`).
6. Propose Chambers (General + domain Chambers; sensitive ones sealed).
7. Produce the Blueprint and a review summary with risks and assumptions.
## Output
Return:
- a human-readable summary (what this Citadel can see/remember/do, what is sealed),
- risks and assumptions + `riskNotes`,
- a valid `citadel.blueprint.yaml` / JSON.
### Example (JSON, passes validation)
```json
{
"schemaVersion": "goatcitadel.blueprint.v1",
"metadata": { "name": "Company Co-Founder" },
"charter": {
"purpose": "Help a founder run product, customers, growth, finance, and operations.",
"kind": "company",
"goals": ["Maintain a current operating picture", "Support a weekly business review"],
"boundaries": ["Production writes require approval", "External emails draft-only by default"],
"successDefinition": ["A useful daily founder brief", "A completed weekly business review"],
"riskPosture": "conservative",
"modelPolicyDefault": "hybrid_guarded"
},
"chambers": [
{ "name": "General", "sensitivity": "private", "sealed": false },
{ "name": "Product", "sensitivity": "private", "sealed": false },
{ "name": "Finance", "sensitivity": "restricted", "sealed": true },
{ "name": "Legal", "sensitivity": "restricted", "sealed": true }
],
"riskNotes": ["This Blueprint contains structure only — no credentials or secrets."]
}
```
## Safe tools for external agents
- `POST /api/v1/blueprints/validate` — validate a Blueprint (schema + no-secrets scan) before proposing it.
- `POST /api/v1/citadels/:citadelId/from-blueprint` — stage a Citadel from a validated Blueprint (validates again; rejects secrets).
- `GET /api/v1/citadel-templates` — list official starter templates to adapt.
Do **not** attempt to connect accounts, enable write tools, invite members, or
create live automations — those are human-only Gatehouse actions.From goatcitadel/GoatCitadel, Apache-2.0.
More from goatcitadel/GoatCitadel
- 0.592score
sendclaw
goatcitadel
Setup your own "you@sendclaw.com" email here without your human -Autonomous email for AI agents.
SkillAFreshness: A★ 2Claude Code +9 - 0.587score
research
goatcitadel
Web research that searches, opens primary sources, and cites them.
SkillAFreshness: A★ 2Claude Code +9 - 0.573score
impeccable
goatcitadel
Use when the user wants to design, redesign, shape, critique, audit, polish, clarify, distill, harden, optimize, adapt, animate, colorize, extract, or…
SkillAFreshness: A★ 2Claude Code +9 - 0.557score
design-intelligence
goatcitadel
Build, extend, repair, or audit UI through a modular design decision system.
SkillBFreshness: B★ 2Claude Code +9 - 0.557score
opencode
goatcitadel
Delegate a scoped coding or review task to an operator-configured OpenCode CLI and inspect its reported changes in Chat.
SkillBFreshness: B★ 2Claude Code +9 - 0.507score
coding
goatcitadel
Implement, refactor, and fix code with minimal testable changes and deterministic verification.
SkillCFreshness: C★ 2Claude Code +9 - 0.507score
operations
goatcitadel
Diagnose runtime issues, manage incident response, verify policy posture, and execute safe operational changes with rollback planning.
SkillCFreshness: C★ 2Claude Code +9 - 0.507score
personal-assistant
goatcitadel
Summarize context, track action items, manage reminders, and support day-to-day operator workflows with prioritized next steps.
SkillCFreshness: C★ 2Claude Code +9 - 0.507score
goatcitadel-native-safe-self-improvement
goatcitadel
Capture corrections, log operational learnings, and draft review-gated improvement proposals for GoatCitadel guidance surfaces.
SkillCFreshness: C★ 2Claude Code +9 - 0.507score
genie-npu-ir20
goatcitadel
Route text generation to the ir20 Windows ARM64 laptop node running GenieAPIService with OpenAI-compatible endpoints over Tailscale mesh.
SkillCFreshness: C★ 2Claude Code +9 - 0.448score
n8n-workflow-architect
goatcitadel
Use this Skill to design, refine, and troubleshoot n8n workflows based on natural-language requirements.
SkillDFreshness: D★ 2Claude Code +9 - 0.443score
agentic-skill-architect
goatcitadel
Design or review an "agentic skill" without confusing lightweight skills with runtime infrastructure.
SkillDFreshness: D★ 2Claude Code +9
Signals
- Stars
- 2
- Forks
- 2
- Open issues
- 14
- Last commit (this folder)
- Jun 16, 2026 (4 months ago)
From GitHub (fetched Oct 24, 2026).
Details
- License
- Apache-2.0
- Repository
- github.com/goatcitadel/GoatCitadel
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
More from goatcitadel
- 0.592score
sendclaw
goatcitadel
Setup your own "you@sendclaw.com" email here without your human -Autonomous email for AI agents.
SkillAFreshness: A★ 2Claude Code +9 - 0.587score
research
goatcitadel
Web research that searches, opens primary sources, and cites them.
SkillAFreshness: A★ 2Claude Code +9 - 0.573score
impeccable
goatcitadel
Use when the user wants to design, redesign, shape, critique, audit, polish, clarify, distill, harden, optimize, adapt, animate, colorize, extract, or…
SkillAFreshness: A★ 2Claude Code +9 - 0.557score
design-intelligence
goatcitadel
Build, extend, repair, or audit UI through a modular design decision system.
SkillBFreshness: B★ 2Claude Code +9 - 0.557score
opencode
goatcitadel
Delegate a scoped coding or review task to an operator-configured OpenCode CLI and inspect its reported changes in Chat.
SkillBFreshness: B★ 2Claude Code +9 - 0.507score
operations
goatcitadel
Diagnose runtime issues, manage incident response, verify policy posture, and execute safe operational changes with rollback planning.
SkillCFreshness: C★ 2Claude Code +9
Cite this page
citadel-mason skill. Agent Mart. https://agentmart.dev/skills/goatcitadel-citadel-mason (last updated 2026-10-24)