sovereign-skills
22 skills for the full Claude Code project lifecycle: setup and task scope, code review, pre-push checks and session notes.
- 151 ★ on the repository
- Last commit
Extended access
Why this level
- The skills create and edit project files (`CLAUDE.md`, rules, hooks) and run scripts.
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add skills/sovereign-skillsDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewExtended accessThe tool runs code or makes network calls and only reads external data.
- The skills create and edit project files (`CLAUDE.md`, rules, hooks) and run scripts.
This is third-party code. Review the repository files before installing.
What it does
A set of 22 standalone skills. Setup: `project-init` builds CLAUDE.md and ROADMAP, `setup` prepares rules, hooks and memory. During work: `scope` fixes what belongs to the task, `freeze` locks the editable zone, `goal-lock` holds the PLAN, DO, VERIFY loop. `pre-push` scans for secrets and runs a review, `code-autopsy` scores code with 12 questions and gives a verdict. Sessions have `session-start` and `session-checkpoint`, audits have `project-check`, `doc-drift` and `eval-leakage-audit`. Each skill is useful alone. Some ship helper Python scripts.
Who it is for. People who run a project in Claude Code for weeks and want to lock in discipline: task boundaries, pre-push checks and context handoff between sessions.
Good fit when
- A new project needs CLAUDE.md, a roadmap and rules in one pass.
- Before a feature, what is in and out of the task must be fixed.
- Every push must pass a secret scan and review.
- A session is ending and the next one must pick up the results.
Not a fit when
- You need a single quick answer or a one-line fix.
- The agent does not read `SKILL.md` Markdown instructions.
- One or two skills of your own are enough and the full set would be excess.
Example request
/scope add CSV report export, then /freeze and implement only inside the reports folder.Limitations
The skills target Claude Code and the shared `SKILL.md` format; for Codex they are copied into `.agents/skills/`. Helper scripts require Python 3.9 or newer, without it a skill falls back to a manual step or stops. Descriptions and triggers are written in English. Many skills share session files such as `tasks/lessons.md` and `~/.claude/STATE.md`, so make sure they do not overwrite yours. The marketplace is third-party and does not auto-update.
How to disable. Delete the chosen skill folders from `~/.claude/skills/` or run `/plugin uninstall` for installed plugins. Created files (`CLAUDE.md`, session notes) remain.
FAQ
Do I need all 22 skills?
No, each works alone. Copy the folder you need into `~/.claude/skills/`, for example `cp -r goal-lock ~/.claude/skills/`.
How do I install through the marketplace?
Run `/plugin marketplace add AlexZio00/sovereign-skills`, then `/plugin install goal-lock@sovereign-plugins`. Third-party marketplaces do not update themselves by default.
Does the set work in Codex?
Yes, copy the skill folder into `.agents/skills/<name>/`. `agents/openai.yaml` is optional and only carries app metadata.
Where to start on a new project?
With `/project-init`, then `/setup`. In daily work `/session-start`, `/scope`, `/pre-push` and `/session-checkpoint` come in handy.
What about an existing project?
Run `/project-check`: it scores infrastructure, security, quality and harness across four dimensions and lists gaps by severity.
Discussion
No questions or reviews yet.
Related
SkillsA Claude Code working configuration: a planner with per-milestone review, sub-agents and skills for analysis, refactoring and docs sync
lencx skills
lencx-skills
Two skills for Claude Code and Codex: keel works through architecture questions, coding-protocol keeps code edits inside the task
A Claude Code environment installer: CLAUDE.md, rules, skills, agents, hooks and two-layer project memory in one script
14 skills that find the gaps in a task before the code: interviews, plans, assumption tests and proof that a bug fix works
Eleven skills for Claude Code and Codex: test audit, git worktree cleanup, plan pages, screen recording transcripts and 1Password sign-in
Standalone skills for the whole development lifecycle: product, architecture, planning, implementation, QA, delivery and operations
README in short
Each skill installs separately, and the whole set covers the cycle from setup to session management and audits.
README badge
Original SKILL.md
--- name: goal-lock description: "Agent Discipline Engine — lock the goal, run PLAN→DO→VERIFY→FINALIZE→OUTPUT loop, detect success masquerading. Triggers: '/goal-lock', '/goal-lock quick', 'goal lock', 'task harness'." user-invocable: true ---