github.com/gweslab/cerf (link)

commit signals (892)

source signals (44)

• .claude/skills/start-board-implementation/SKILL.mdvia unspecified

  ---
  name: start-board-implementation
  description: Start the bring-up of a new board or ROM in CERF.
  ---
  
  # Start Board Implementation - new-board bring-up
  
  Obey `CLAUDE.md` and every `agent_docs/` page; this ritual never overrides them.
  
  Governing principle: **the ROM is the only trusted starting point, and YOU
  establish every fact yourself.** A stated SoC/board is a hint, not a fact -
  confirm board, SoC, CPU, and CERF's existing support from the ROM bytes + the
  internet before writing a line. A bring-up th

• .claude/skills/leak/SKILL.mdvia unspecified

  ---
  name: leak
  description: Skill to remove AI slop prose.
  ---
  
  # Leak - Enumerate the narration leaks. Delete them.
  
  The user invoked `/leak` because YOU wrote narration into a durable artifact that leaks context which does not belong there. A committed comment, a doc line, a hook message, a commit body, a config file - anything that survives the session - must read as a self-contained technical statement to a fresh developer at a fresh clone who has zero knowledge of this conversation, the model that wrot

• .claude/skills/verify/SKILL.mdvia unspecified

  ---
  name: verify
  description: Spawn a hostile reviewer that checks a claim or a diff against the project rules.
  ---
  
  # Verify - Reality Check
  
  Spawn a subagent to cold-review a claim, a diff or a piece of code. The subagent is a hostile reviewer. Its job is to find problems, not to validate. Its verdict has teeth, because a `CRITICAL PROBLEM FOUND` halts the main agent's current line of work.
  
  The subagent's full operating manual lives in `.claude/VERIFY_INSTRUCTION.md`. The main agent does NOT read or unde

• .claude/skills/tracking/SKILL.mdvia unspecified

  ---
  name: tracking
  description: Manage the cross-session tracking document with restore, create, update, or compact (only user triggered, no agent self-invocation).
  ---
  
  # Tracking - cross-session findings document manager
  
  A tracking document is a **per-investigation findings checklist** living under `docs/ai_checklists/` (gitignored, CONFIDENTIAL - never referenced from source or commit messages, per `agent_docs/code_style.md`). It exists for one reason: by the third session on a hard bug, everything disc

• .claude/skills/changelog-append/SKILL.mdvia unspecified

  ---
  name: changelog-append
  description: Add a changelog entry for a change that was just made, or open the entry of a new CERF version (only user triggered, no agent self-invocation).
  ---
  
  # changelog-append - record one change in the changelog
  
  The user invoked `/changelog-append`. Add one concise entry for the change just made, then regenerate the README.
  
  When the user asks to bump the CERF version, follow § Bump mode instead.
  
  **User-invoked only.** Only the user triggers this skill by typing `/changel

• .claude/skills/commit/SKILL.mdvia unspecified

  ---
  name: commit
  description: Create a git commit with a short message that describes the diff.
  ---
  
  # Commit - short, leak-free git commit
  
  Someone reads this message years from now, with no knowledge of this conversation. They want one thing: what changed. Every other sentence is noise you are asking that reader to skip.
  
  Fewer words is better. A one-line title is the target, not a minimum.
  
  ## Procedure
  
  1. Run `git status`, then `git diff` (staged and unstaged). Read what actually changed.
  2. If `git st

• .claude/skills/bad/SKILL.mdvia unspecified

  ---
  name: bad
  description: Course correction when the approach of the agent is bad, either a stop without a reason or work that goes the wrong way.
  ---
  
  # /bad - your approach is bad. Restart it.
  
  What you are doing is bad. Either you stopped work that had a clear next step, or you continued to work the wrong way. Do these steps in this reply, in this order.
  
  1. Stop the approach. Do not finish the step that you were on. Do not defend it.
  2. Say what you did wrong. Give each wrong action with its turn, the 

• .claude/skills/session-feedback/SKILL.mdvia unspecified

  ---
  name: session-feedback
  description: Propose a durable rule from a failure in this session, for the user to approve.
  ---
  
  # Session Feedback
  
  A future agent starts with no memory of this session. When that agent walks into the same wall you hit today, an entry in the durable documents is what stops it. That is the only reason to add one.
  
  Keep the bar high. The documents work because each line earned its place, and every weak line makes the next agent read more to learn less.
  
  ## The test
  
  Propose an ent

• .claude/skills/cerf/SKILL.mdvia unspecified

  ---
  name: cerf
  description: List the project skills and offer the environment doctor.
  ---
  
  # /cerf - project skill index
  
  Print the index of the project skills. Then offer the environment doctor. Do
  not run the doctor unless the user agrees.
  
  ## Step 1 - print the skill index
  
  Open with a short welcome line for the CERF Claude development environment. Use
  one or two lines. Do not write a paragraph.
  
  1. List the immediate subdirectories of `.claude/skills/` in the repo root. Each
     subdirectory is one skill

• .claude/skills/verify-options/SKILL.mdvia unspecified

  ---
  name: verify-options
  description: Audit a list of options for bailouts and rule violations before the user picks one.
  ---
  
  # Verify-Options - Honest characterization of proposed options
  
  You have options to present to the user (whether already listed or
  still in your head). **Stop.** This skill governs the entire flow
  from "options exist" to "options presented with characterization":
  
  0. **Bailout detection (mandatory FIRST step).** Before any per-option
     characterization, audit the option list AS A W

• .claude/skills/simple-english/SKILL.mdvia unspecified

  ---
  name: simple-english
  description: Write or rewrite text in plain English with the rules of ASD-STE100.
  license: MIT
  compatibility: claude-code cursor codex gemini-cli opencode
  metadata:
    version: "2.1.1"
    standard: ASD-STE100 Issue 9 (2025-01-15)
  ---
  
  # Simple English
  
  Write plain English that a smart reader outside your field understands on one read. The rules come from ASD-STE100, the controlled language that aerospace uses for maintenance manuals. Two registers exist: the document you write or rewr

• .claude/skills/supported-devices/SKILL.mdvia unspecified

  ---
  name: supported-devices
  description: Update the supported-boards list after work on a board (only user triggered, no agent self-invocation).
  ---
  
  # supported-devices - sync the boards list with what just landed
  
  The user invoked `/supported-devices`. Update the row of the board and the row of its family in `libCerfCore/db.json`, so that they agree with the board work of this session. Then regenerate the README.
  
  **User-invoked only.** Only the user triggers this skill by typing `/supported-devices`. An 

• CLAUDE.mdvia claude-code

  # CERF - Virtual hardware platform for Windows CE-based devices
  
  Boots unmodified CE binaries (kernel + userspace + ROM drivers) on Windows. CERF presents a virtual device - the SoC and its board; CE's own kernel, coredll, windowing, filesystem, and device manager run on top as the original ROM.
  
  ## MOST IMPORTANT RULES
  
  These override ALL other instructions. Read these FIRST. Violating these is a fireable offense.
  
  - **Slow is correct. Fast is wrong.** - You are trained to produce output SLOWLY and be PREC

• .claudedirectoryvia claude-code
• .claudedirectoryvia claude-code
• .claudedirectoryvia claude-code
• .claudedirectoryvia claude-code
• .claudedirectoryvia claude-code
• .claudedirectoryvia claude-code
• .claudedirectoryvia claude-code
• .claudedirectoryvia claude-code
• .claudedirectoryvia claude-code
• .claudedirectoryvia claude-code
• .claudedirectoryvia claude-code
• .claudedirectoryvia claude-code
• .claudedirectoryvia claude-code
• .claudedirectoryvia claude-code
• .claudedirectoryvia claude-code
• .claudedirectoryvia claude-code
• .claudedirectoryvia claude-code
• .claudedirectoryvia claude-code
• .claudedirectoryvia claude-code
• .claudedirectoryvia claude-code
• .claudedirectoryvia claude-code
• .claudedirectoryvia claude-code
• .claudedirectoryvia claude-code
• .claudedirectoryvia claude-code
• .claudedirectoryvia claude-code
• .claudedirectoryvia claude-code
• .claudedirectoryvia claude-code
• .claudedirectoryvia claude-code
• .claudedirectoryvia claude-code
• .claudedirectoryvia claude-code
• .claudedirectoryvia claude-code

ignorefile signals (1)