ADLC Team Skills
Skills that give a coding agent your team rules at session start: a rule index, a task factory, decision records and evals
- 145 ★ on the repository
- Last commit
- MIT
Extended access
Why this level
- The skills publish decision records back to the shared team repository, so a mistake reaches every member
- The `session_start` handler changes the context of every session, and gates and logs add steps to tasks
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add skills/adlc-team-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 publish decision records back to the shared team repository, so a mistake reaches every member
- The `session_start` handler changes the context of every session, and gates and logs add steps to tasks
This is third-party code. Review the repository files before installing.
What it does
ADLC Team Skills exist so that an agent does not work like a stranger on a team. When a session opens, the `team-boot` skill fires on the `session_start` event and injects a lean index of team rules: about a hundred tokens of names and one-line descriptions, without the rules themselves. When a task matches a rule, the agent pulls its full text. Rules live in the `team-ai-directives` git repository reviewed by pull requests. The `factory-mission` skill keeps the agent from jumping to code: first a contract is fixed (goal, constraints, non-goals, success criteria), then comes a chain of specify, plan, implement and converge with gates, a circuit breaker, resume and an audit trail. The `team-levelup` skill turns a fix found in a session into a Context Directive Record (CDR), scores its confidence and publishes accepted records back to the team repository. Usage data lives in an orphan branch `adlc`. The `evals-*` skills build executable checks on promptfoo with binary graders. The repository also holds the `team-init`, `team-learn` and `team-repair` skills and documents on decision memory and spec-kit integration.
Who it is for. Development teams where several people work with agents and rules sit in a shared repository instead of personal CLAUDE.md files, and tech leads responsible for one process.
Good fit when
- Team rules must reach the agent at every start without bloating the context
- The agent must agree on a task contract before writing code
- Found fixes have to be saved as records and published for the whole team
- The behaviour of rules and skills must be checked by executable graders
Not a fit when
- A single developer on a single project: a team rules layer is excessive
- There is no team rules repository (`team-ai-directives`) and none is planned
- You need a manual process with no gates, circuit breakers or audit trail
Example request
Start the task through factory-mission: first draft a short contract with the goal, constraints and success criteria and show it to meLimitations
The skills assume a team with a shared `team-ai-directives` rules repository and agents with session events. They add checks, gates and logs, so small tasks take more steps. Usage data is kept in the `adlc` branch of the repository. Evals use promptfoo and need separate setup. The last commit is dated November 10, 2026. The exact install command is in the README Install section and is worth reading before running.
How to disable. Delete the installed `team-*`, `factory-*` and `evals-*` skills from the agent skills directory and remove the `session_start` event handler that wires them up.
FAQ
What does team-boot do?
It fires at session start and injects a lean team rule index of about a hundred tokens, and the full text of a rule loads when the task matches it.
What is factory-mission?
A skill that fixes the task contract first and then walks it through specify, plan, implement and converge with gates and an audit trail.
What is a CDR?
A Context Directive Record: a fix found in a session. The `team-levelup` skill scores its confidence and publishes accepted records to the team repository.
Where is usage data stored?
In the orphan `adlc` branch of the repository.
How are the skills themselves verified?
With the `evals-*` skills, which build executable checks on promptfoo with binary graders.
Discussion
No questions or reviews yet.
Related
SkillsADLC is a set of dependency-free CLIs and agent plugins: gates at each development phase against typical model failure modes.
A set of independent skills against drift in large codebases: hard-cut, root-cause tracing, lean testing and CI right-sizing.
A library of 30-plus instruction skills for coding agents: planning, building, testing, hallucination guards and shipping
Nine short skills: tests, review, ADR plans, slides, a wiki, image generation, worktrees and plan implementation.
Skill and plugin for Codex and Claude Code: the agent drives a project through a plan-to-maintain loop with human gates.
Two skills for Codex and Claude Code: prd keeps a living implementation checklist and review audits designs from the user's side
Comparison with similar tools
Cards that share the main search query, and how they differ.
README in short
The README explains the idea: team rules reach the agent at session start and tasks pass through a contract and gates. It continues with sections on the developer journey, the problems solved, installation, the task factory, universal orchestration, contents and philosophy, plus security and the release process.