菜单
Manage experimental spike projects in a centralized, searchable location. Use when you need to create, find, or navigate to spike/experiment directories.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
| name | spiker |
| description | Manage experimental spike projects in a centralized, searchable location. Use when you need to create, find, or navigate to spike/experiment directories. |
| allowed-tools | Bash(guppi-spiker:*) |
| version | 0.4.1 |
| author | Sam Dengler |
| license | MIT |
Organizes experimental coding projects into a single directory with date-prefixed naming, search, metadata tracking, and optional git init. Creates directories like 2026-02-12-redis-caching under a configurable root. Optionally tracks summaries, tags, and lifecycle status via beads.
guppi-spiker new [name]Create a new spike directory.
Arguments:
name (optional) — slug for the directory. If omitted, generates a random name (e.g., fuzzy-teal-otter)Options:
--no-git — skip git init--summary / -s — one-line summary (stored in beads)Examples:
guppi-spiker new redis-caching # ~/spikes/2026-02-12-redis-caching/
guppi-spiker new # ~/spikes/2026-02-12-fuzzy-teal-otter/
guppi-spiker new redis-test --summary "Testing pub/sub for notifications"
guppi-spiker list [query]List all spikes with summaries, most recent first. Optional query filters by slug substring.
Options:
--all / -a — include done (closed) spikes--status — filter by status (open, in_progress, deferred, closed)guppi-spiker list
guppi-spiker list redis
guppi-spiker list --all
guppi-spiker list --status deferred
guppi-spiker go <query>Print the path to the most recent matching spike. Use with cd $(...).
cd $(guppi-spiker go redis)
guppi-spiker describe <query> <summary>Set or update a spike's summary.
guppi-spiker describe redis "Testing pub/sub for notifications"
guppi-spiker tag <query> <tags...>Add tags to a spike.
guppi-spiker tag redis backend caching
guppi-spiker park <query>Park a spike (mark as deferred — paused but still visible in list).
guppi-spiker park redis
guppi-spiker done <query>Mark a spike as done (closes the beads issue, hidden from default list).
guppi-spiker done redis
Set SPIKER_PATH env var to change the spike root directory (default: ~/spikes).
guppi-spiker skill install # Register with guppi-cli
guppi-spiker skill show # Display SKILL.md contents
Rank movie showtimes by the user's format preferences and find the best contiguous seats by their saved seat preferences. Use when the user wants to find movie tickets/seats (currently Regal/regmovies.com). Booking is human-in-the-loop — usher proposes seats and a deep link; the user logs in and pays.
Analyze academic papers using the Feynman Technique. Use when you need to create a clear, structured explanation of an academic paper.
Cross-project task and idea tracker built on beads. Use when you need to capture ideas, tasks, reading lists, or track async work across projects.
Research historical events from local history sources (Chrome, terminal). Use when you need to find what happened, when it happened, or retrace activity.
Telegram-based messaging for Claude workflows. Use when you need to send or receive messages and files between Claude Desktop and Claude Code.
Deterministic secret storage for guppi skills. Use when you need to store, retrieve, or manage API tokens and secrets across guppi skills.