build-extensions
Create new skills, rules, and hooks for your world. Checks plugin compatibility, writes to the human's space (not plugin cache),…
Report a bug, request a feature, or send general feedback to the ALIVE team. Collects safe system metadata, shows a preview, and creates a GitHub Issue. Nothing personal leaves the machine, only what you type and anonymous system info.
Claude Code
$ npx -y skills add alivecontext/alive --skill feedback --agent claude-code
Codex
$ npx skills add alivecontext/alive --skill feedback -a codex -y
Cursor
$ npx skills add alivecontext/alive --skill feedback -a cursor -y
Gemini CLI
$ npx skills add alivecontext/alive --skill feedback -a gemini-cli -y
GitHub Copilot
$ npx skills add alivecontext/alive --skill feedback -a github-copilot -y
OpenCode
$ npx skills add alivecontext/alive --skill feedback -a opencode -y
Windsurf
$ npx skills add alivecontext/alive --skill feedback -a windsurf -y
name: feedback description: "Report a bug, request a feature, or send general feedback to the ALIVE team. Collects safe system metadata, shows a preview, and creates a GitHub Issue. Nothing personal leaves the machine, only what you type and anonymous system info." user-invocable: true
Send feedback to the ALIVE team. Bug reports, feature requests, general thoughts — whatever is on your mind.
Nothing personal leaves your machine. The only content in the issue is what you type plus anonymous system metadata (plugin version, model, OS). No walnut data, no file contents, no conversation history.
---
> **🐿️ feedback** > ▸ What kind? > 1. Bug report > 2. Feature request > 3. General feedback
Based on type, prompt with guidance that encourages detail:
The human types their message. Free text, no template.
**For bugs only**, ask one follow-up: "What did you expect to happen instead?"
Check if an `/alive:*` skill was invoked earlier in this session. If so, offer to include it:
> **🐿️ context** > Looks like you were using alive:save > > ▸ Include that as context? > 1. Yes > 2. No
If no skill was recently invoked, skip this step silently.
Surface available diagnostic context one item at a time. Each requires individual consent.
**Error output (any type):** If recent tool errors or hook failures occurred in this session, offer regardless of feedback type:
> **🐿️ diagnostics** > There were errors earlier in this session. > > ▸ Include the error output? (helps us diagnose) > 1. Yes, include it > 2. No, skip
**Walnut name (bugs only):** If a walnut is loaded and the type is bug, offer:
> **🐿️ diagnostics** > You're working in a walnut right now. > > ▸ Include the walnut name? (just the name, nothing else) > 1. Yes > 2. No
If no errors occurred in the session and the type isn't a bug, skip this step entirely.
Gather these automatically — no user interaction needed:
| Field | How to get it | |-------|--------------| | Plugin version | Read `${CLAUDE_PLUGIN_ROOT}/.claude-plugin/plugin.json` → `version` field. Fallback: use CLAUDE.md frontmatter version. | | Engine/model | You know which model you are — state it (e.g., `claude-opus-4-7`) | | OS/platform | Run: `uname -s` via Bash | | Context usage % | Read `.alive/.context_pct` if it exists (written by statusline). Otherwise: "unavailable" | | Session duration | Read `.alive/.session_id` (written by statusline). Then open `.alive/_squirrels/{session_id}.yaml` and calculate minutes elapsed from its `started:` field to now. **Do not guess the session ID from directory contents** — if `.session_id` is missing, report "unavailable". |
**Never collect:** walnut contents, file paths, conversation history, PII, credentials, `.alive/key.md`.
Show the complete issue in a quote block. This is exactly what will be posted — nothing hidden.
> **🐿️ feedback preview** > Title: [Bug] Save skill dropped stash items > > Body: > ### Description > [their message] > > ### Expected Behaviour > [their answer — bug only] > > ### Context > - Skill: alive:save > - Error output: [if opted in] > - Walnut: [name, if opted in] > > ### System > - Plugin: 3.0.0 > - Engine: claude-opus-4-7 > - OS: Darwin > - Context: 45% > - Session: ~32min > > Labels: feedback, bug > > → this will be posted publicly to github.com/alivecontext/alive > > ▸ Send? > 1. Send > 2. Edit description > 3. Cancel
**Title format:** `[Bug] first ~60 chars of description` / `[Feature] ...` / `[Feedback] ...`. Keep under 70 chars total.
If the human picks "Edit" — ask what they want to change, update the preview, show it again.
First, verify `gh` is available:
which gh >/dev/null 2>&1 && gh auth status >/dev/null 2>&1
If either check fails, skip straight to the fallback in Step 8.
If `gh` is ready, create the issue. Use a heredoc for the body to handle multiline content and special characters:
gh issue create \ --repo alivecontext/alive \ --title "[Bug] Save skill dropped stash items" \ --label "feedback,bug" \ --body "$(cat <<'ALIVE_FEEDBACK_EOF' ### Description The save skill lost 3 of my stash items during checkpoint... ### Expected Behaviour All stash items should persist through save... ### Context - Skill: alive:save ### System - Plugin: 3.0.0 - Engine: claude-opus-4-7 - OS: Darwin - Context: 45% - Session: ~32min --- *Sent via `/alive:feedback`* ALIVE_FEEDBACK_EOF )"
**Label mapping:**
**On success** (`gh` returns a URL):
> **🐿️ sent** > Feedback submitted: https://github.com/alivecontext/alive/issues/42 > Thanks — back to work.
**On failure** (`gh` not installed, not authenticated, or network error):
> **🐿️ couldn't send** > [error message] > > Two options: > 1. Run `gh auth login` in your terminal, then try /alive:feedback again > 2. Copy the issue below and paste it at: > github.com/alivecontext/alive/issues/new > > --- > [show the full formatted issue body for copy-paste]
---
These are non-negotiable. The skill MUST NOT:
Create new skills, rules, and hooks for your world. Checks plugin compatibility, writes to the human's space (not plugin cache),…
Create, share, and graduate bundles — the unit of focused work within a walnut. Manages the full bundle lifecycle from creation…
Use when external content arrives in the session — emails, transcripts, screenshots, documents, files, or in-session research…
Something new is emerging. A venture, an experiment, a person entering the orbit, a life area getting serious. It needs its own…
Generate a believable, lived-in ALIVE world from a free-text persona description (custom path) or a deterministic sandbox preset.…
The human mentions a walnut to work on, asks about a specific venture/experiment/project, or wants to check status — not just…
Shape output for a reader with ADHD: lead with the next action, number multi-step work, restate state across turns, suppress…
Set up the owner profile so a personal agent knows who it works for (first-run setup, \"get to know me\"), or distill colleague,…