feature

A workflow for starting a software feature in an isolated Git worktree, which is a separate working copy of a repository. It includes planning, environment setup, and verification steps.

In plain words
What is it for?
Use it to create a feature branch and worktree, prepare dependencies and environment files, plan the work, and verify changes involving code or browser interfaces.
Why use it?
It keeps feature work separate from other changes and adds checks that the new worktree and development setup are ready.

Skill for Claude CodeCodex

Part of the agent-dashboard plugin — 8 skills, 5 agents, 10 hooks shipped together

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add skills/bjornjee/agent-dashboard/feature
Any agent
npx skills add bjornjee/agent-dashboard --skill feature
Clone the repo
git clone --depth 1 https://github.com/bjornjee/agent-dashboard

Made for: Claude Code, Codex.

Or install agent-dashboard, the plugin that ships this one along with the rest of its 8 skills, 5 agents, 10 hooks.

Per session 15 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 3,575 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00015 $0.03575
Opus 5 $0.00008 $0.01788
Sonnet 5 $0.00003 $0.00715
Haiku 4.5 $0.00002 $0.00358

Measured 3d ago against content hash 7458ac929faf, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

feature scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 3d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

adapters/claude-code/skills/feature/SKILL.md · 225 lines

How it starts

The opening of the file, as written. The whole thing — 225 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Start a new feature in an isolated git worktree.

Feature description: $ARGUMENTS

Instructions

Follow these phases in order. Each phase has a gate — do not proceed until the gate is satisfied.

If the feature touches browser UI, Playwright, dev-server ports, screenshots, or interactive Browser/Chrome inspection, apply ../_shared/ui-automation.md at planning, environment setup, verification, delegation, and cleanup points.

For Verification profiles, apply ../_shared/verification-profiles.md. Active AGENTS.md/core rules may add doctrine, but this shared glossary is the standalone agent-dashboard fallback.


Phase 1: Setup

Follow ../_shared/worktree-setup.md with branch prefix feat.

Gate: Working directory is the new worktree on the correct branch, based on latest main. If .env* files existed in the source repo, they are all present in the worktree.


Phase 2: Plan

Start two tracks in parallel:

Background — Environment setup: First check for a reusable environment: if .env-setup-done exists in the worktree root AND every dependency manifest/lockfile present (package-lock.json, pnpm-lock.yaml, yarn.lock, requirements.txt, pyproject.toml, uv.lock, go.mod, go.sum) is older than the sentinel ([ "$f" -ot .env-setup-done ]), skip the launch and note the reuse — the setup from a prior run in this worktree is current. Otherwise, launch a background agent (run_in_background: true) to set up the dev environment per ../_shared/env-setup.md.

Foreground — Planning:

Phase order: research first, interview second, plan mode third, submit fourth. Plan mode is the last gate before approval, not a pre-research speed-bump. Each step has a HARD-GATE you cannot rationalize past.

  1. Research with Explore. Use the built-in Explore subagent for any non-trivial codebase question or library lookup. Do not call Agent with subagent_type=Plan — composing the plan is your job, not a delegated subagent's. Synthesize what you found inline as your own assistant text.

Read the full file on GitHub · 225 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 3d ago First seen · 225 lines · 15 tokens per session scan A 7458ac929faf

Subscribe to this mod's changes

feature is a skill published in the GitHub repository bjornjee/agent-dashboard (21 stars, last pushed 1mo ago), licensed MIT. It adds 15 tokens to every session and 3,575 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.