PR

project-reentry-guide

Creates a canonical entry document to guide developers during project onboarding.

Install

mkdir -p .claude/skills/project-reentry-guide && curl -L -o skill.zip "https://agentskills.codes/api/skills/download/18352" && unzip -o skill.zip -d .claude/skills/project-reentry-guide && rm skill.zip

Installs to .claude/skills/project-reentry-guide

Activation

This is the description your AI agent reads to decide when to run this skill — the better it matches your request, the more reliably it fires.

Create/update docs/design/START_HERE.md so onboarding and re-entry always start from the same place.
100 charsno explicit “when” trigger
Beginner

Key capabilities

  • Create/update `docs/design/START_HERE.md`
  • Include top canonical documents
  • Map project domain with key components
  • List important entry files for the codebase
  • Provide a checklist of re-entry invariants
  • Summarize active TODOs

How it works

This skill generates or updates a `START_HERE.md` file by compiling essential project information, including canonical documents, domain map, entry points, and invariants, to aid onboarding and re-entry.

Inputs & outputs

You give it
Existing documentation, current TODOs, repository tree
You get back
Structured `docs/design/START_HERE.md` file

When to use project-reentry-guide

  • Standardize project onboarding
  • Create entry documentation
  • Map project domain

About this skill

project-reentry-guide

Canonical docs (Top 5)

  1. semantic-cad-architecture.md — coordinates/units/grid/UX philosophy (primary)
  2. architecture-overview.md — layers, APIs, HTTP schema
  3. state-transitions.md — strict state machine + dirty resolution rules
  4. floorplan-workspace-strategy.md — floorplan workspace design
  5. wooden-frame-preset-plan.md — SRBA A-F + wooden preset plan

Domain map (TOP 5 + alpha)

  • kernel: src/kernel/ — exact geometry kernel (Point2/Line2/Circle2/Rect2), f64 correctness
  • scene: src/scene/ — SceneContext/SceneWorld, ECS, dirty flags, GPU sync
  • floorplan: src/floorplan/ — semantics layer (Space/Zone/Level), wall gen, resolution
  • fsm: src/fsm/ — command FSM (Line/Circle/Rect/Move/Rotate/Scale/Space)
  • renderer: src/renderer/ — wgpu pipeline, frame generation
  • ui: src/ui/ — egui UI (menus, property panel)
  • server: src/server/ — HTTP+JSON agent control (screenshot/state)

Entry files (Top 5)

  • src/main.rs — event loop, winit/egui integration, startup sequence
  • src/context.rs — current SceneContext API implementation
  • src/scene/context.rs — new SceneContext (design-doc aligned; future source of truth)
  • src/floorplan/mod.rs — FloorplanModel (Space/Zone/Level)
  • src/app.rs — App FSM, Msg/update, tool switching, global state

Re-entry invariants (checklist)

  • RH coords: X=East, Y=North, Z=Up, 1 unit = 1mm (semantic-cad-architecture.md §1)
  • Scale is View; Model is always mm (semantic-cad-architecture.md §2)
  • SceneContext is the only public API; do not touch ECS directly (architecture-overview.md, AGENTS.md)
  • State rules: Display/Select/Highlight; Visible=false select/highlight is error (state-transitions.md)
  • Dirty resolution order: Geometry -> Transform -> Visual (state-transitions.md)
  • FSM split: System(Input)/App(Mode)/Command(Logic) (AGENTS.md)
  • Dual rep: Kernel(f64) is truth; Mesh is disposable cache (architecture-overview.md)
  • Design docs are authoritative; if code contradicts docs, update docs first (AGENTS.md, agent_checklist.md)
  • SRBA layers A-F (wooden-frame-preset-plan.md)
  • Try-lock policy for CursorMoved; blocking for important events (architecture-overview.md)

Active TODOs (Top 10)

  1. Space/Room input tool — src/fsm/space_rect.rs, src/fsm/space_poly.rs, floorplan-workspace-strategy.md
  2. Floorplan palette UI — src/ui/, floorplan-workspace-strategy.md
  3. SceneContext migration — src/context.rs -> src/scene/context.rs, SCENECONTEXT_MIGRATION.md
  4. AxiomModel new (A layer) — src/axiom/ (planned), wooden-frame-preset-plan.md
  5. WallComposition separation — src/composition/ (planned), wooden-frame-preset-plan.md
  6. DispositionRules — src/disposition/ (planned), wooden-frame-preset-plan.md
  7. Undo/Redo expansion — src/context.rs, addendum-03-persistence-and-undo.md
  8. Drawing View — src/projection/ (WIP), projection_layer.md
  9. Warning cleanup — whole project, AGENTS.md
  10. Semantic Grid — src/grid.rs, semantic-cad-architecture.md

Output file requirements: docs/design/START_HERE.md

Write/update these sections in order:

  1. TL;DR (5 lines)
  2. Current Focus / Last Good State / Next Action (3 short lines)
  3. What to read first (Top docs)
  4. Domain map
  5. Entry files
  6. Active TODOs
  7. Invariants checklist
  8. Start in 5 steps
  9. Missing/unclear items

When not to use it

  • The project does not require onboarding documentation
  • The project does not need a re-entry guide
  • The project is not a software development project

Limitations

  • Keep within 1-2 pages equivalent.
  • Must include: Top docs, domain map, entry files, active TODOs, invariants checklist, and "Start in 5 steps".
  • If something is missing, mark as MISSING and suggest where to add it.

How it compares

This skill standardizes the onboarding process by creating a single, complete `START_HERE.md` file, ensuring new team members always begin with a consistent set of critical project information.

Compared to similar skills

project-reentry-guide side by side with the closest alternatives in the catalog.

SkillInstallsUpdatedSafetyDifficulty
project-reentry-guide (this skill)06moNo flagsBeginner
session-handoff132moReviewBeginner
spec-to-backlog83moNo flagsIntermediate
create-handoff36moReviewIntermediate

Try saying

Example prompts that trigger this skill in your AI assistant.

You might also like

session-handoff

davila7

Creates comprehensive handoff documents for seamless AI agent session transfers. Triggered when: (1) user requests handoff/memory/context save, (2) context window approaches capacity, (3) major task milestone completed, (4) work session ending, (5) user says 'save state', 'create handoff', 'I need to pause', 'context is getting full', (6) resuming work with 'load handoff', 'resume from', 'continue where we left off'. Proactively suggests handoffs after substantial work (multiple file edits, complex debugging, architecture decisions). Solves long-running agent context exhaustion by enabling fresh agents to continue with zero ambiguity.

1370

spec-to-backlog

atlassian

Automatically convert Confluence specification documents into structured Jira backlogs with Epics and implementation tickets. When Claude needs to: (1) Create Jira tickets from a Confluence page, (2) Generate a backlog from a specification, (3) Break down a spec into implementation tasks, or (4) Convert requirements into Jira issues. Handles reading Confluence pages, analyzing specifications, creating Epics with proper structure, and generating detailed implementation tickets linked to the Epic.

868

create-handoff

parcadei

Create handoff document for transferring work to another session

331

moai-workflow-project

modu-ai

Integrated project management system with documentation, language initialization, and template optimization modules. Use when setting up projects, generating documentation, configuring multilingual support, or optimizing templates.

29

granola-core-workflow-b

jeremylongshore

Post-meeting note processing and sharing workflow with Granola. Use when reviewing meeting notes, sharing with team members, or processing action items after meetings. Trigger with phrases like "granola post meeting", "share granola notes", "granola follow up", "process meeting notes", "granola action items".

00

reporting

comeredon

Final pipeline execution report format. Defines the structure, agent registry, Mermaid gantt chart, and change summary. Used by AnalystAgent in Phase 6.

00

Search skills

Search the agent skills registry