Implement Incremental
Runs an implementation plan phase by phase, delegating each phase to specialist agents
Test report
- Verdict
- Works with setup
- Score
- Tested
- Jul 31, 2026
- Environment
- Claude Code 2.x (agent harness)
- Upstream re-checked
- Aug 10, 2026 · 05aa396
Fetched SKILL.md and its three referenced files (reference/output-format.md, examples/output-example.md, reference/perspectives.md) — all HTTP 200; frontmatter parses with name+description; no security smells. Downloaded the repo's own eval-2-resume fixture (phase-1 completed, phase-2 in_progress with task T2.1 already checked, phase-3 pending) and ran the skill's Plan Discovery + resume logic against it, producing baseline.md vs skill.md. Concrete diff: the no-skill baseline re-implements T2.1 (the already-completed controller scaffold) and writes code in the orchestrator, while the skill-following artifact skips phase-1, resumes phase-2, does NOT re-delegate the completed T2.1, delegates only remaining tasks to a subagent, and flips both phase-2 frontmatter status and the README checkbox. Could not execute the actual subagent delegation or the validate/specify-meta sibling-skill calls — those need the full plugin ecosystem, hence setup.
Scored on four weighted criteria — install, triggering, output vs. baseline, docs. How scoring works
- Installs cleanly 5/5
- Triggers reliably 4/5
- Output vs. baseline 6/10
- Docs & honesty 4/5
What Implement Incremental does
A human-in-the-loop orchestrator that executes a scaffolded implementation plan (plan/README.md manifest plus phase-N.md files) one phase at a time, delegating all coding to specialist subagents rather than writing code itself. It skips already-completed phases, resumes in_progress ones, runs a drift/constitution check via the sibling validate skill at each phase boundary, and pauses for user confirmation between phases. Triggers on requests to implement a multi-phase feature plan with review gates; it is user-invocable:false and normally dispatched by the repo's implement router skill.
How to install Implement Incremental
git clone --depth 1 https://github.com/rsmdt/the-startup.git /tmp/implement-incremental-src
mkdir -p ~/.claude/skills
cp -R /tmp/implement-incremental-src/plugins/start/skills/implement-incremental ~/.claude/skills/implement-incremental
# This skill is ONE stage of the "start" plugin pipeline. Standalone it is limited:
# its body calls sibling skills (validate, specify-meta) and expects a plan/ directory
# scaffolded by the specify-incremental skill. For full function install the whole plugin:
# /plugin marketplace add rsmdt/the-startup then /plugin install start
# Note: SKILL.md sets user-invocable: false — it is meant to be dispatched by the
# `implement` router skill (which auto-detects tier), not invoked directly by the user.
Skills live in ~/.claude/skills/ (global) or .claude/skills/
(per-project). Restart Claude Code after installing.
Commands — how to trigger Implement Incremental
-
/implement-incrementalRuns an implementation plan phase by phase, delegating each phase to specialist agents
It also activates on plain-language prompts like these:
-
Implement this feature plan phase by phase -
Review each phase before moving to the next -
Walk through this single-feature implementation plan
Frequently asked questions
- Is the Implement Incremental skill free?
- Yes. The skill itself is free from rsmdt/the-startup. SkillProof publishes the install command and an independent test verdict at no cost.
- Does Implement Incremental work with Claude Code?
- We tested it with Claude Code 2.x (agent harness) on Jul 31, 2026. Verdict: Works with setup. Fetched SKILL.md and its three referenced files (reference/output-format.md, examples/output-example.md, reference/perspectives.md) — all HTTP 200; frontmatter parses with name+description; no security smells. Downloaded the repo's own eval-2-resume fixture (phase-1 completed, phase-2 in_progress with task T2.1 already checked, phase-3 pending) and ran the skill's Plan Discovery + resume logic against it, producing baseline.md vs skill.md. Concrete diff: the no-skill baseline re-implements T2.1 (the already-completed controller scaffold) and writes code in the orchestrator, while the skill-following artifact skips phase-1, resumes phase-2, does NOT re-delegate the completed T2.1, delegates only remaining tasks to a subagent, and flips both phase-2 frontmatter status and the README checkbox. Could not execute the actual subagent delegation or the validate/specify-meta sibling-skill calls — those need the full plugin ecosystem, hence setup.
- What is the Implement Incremental SkillProof Score?
- 7.6/10 — installs cleanly 5/5, triggers reliably 4/5, output vs. baseline 6/10, docs & honesty 4/5.
- How do I install Implement Incremental?
- Copy the install command from this page, run it in your terminal, and restart Claude Code. Skills live in ~/.claude/skills/ (global) or .claude/skills/ inside a project.
- Can I use Implement Incremental with Cursor, Copilot, Gemini CLI, Codex or other AI tools?
- The SKILL.md format is native to Claude (Claude Code, Desktop, claude.ai). The instructions inside adapt to other assistants: Cursor rules, GitHub Copilot instructions, Windsurf rules, Custom GPTs, AGENTS.md for OpenAI Codex, and GEMINI.md for Google Gemini CLI — our conversion guides cover each, and the free converter on the tools page does the wrapping for you.