Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add mikestangdevs/craft-skills --skill checkpoint-handoffgit clone --depth 1 https://github.com/mikestangdevs/craft-skillsWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/mikestangdevs/craft-skills/checkpoint-handoff)<a href="https://agentmods.dev/skills/mikestangdevs/craft-skills/checkpoint-handoff"><img src="https://agentmods.dev/badge/skills/mikestangdevs/craft-skills/checkpoint-handoff/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/mikestangdevs/craft-skills/checkpoint-handoff"><img src="https://agentmods.dev/badge/skills/mikestangdevs/craft-skills/checkpoint-handoff.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00151 | $0.01277 |
| Opus 5 | $0.00076 | $0.00639 |
| Sonnet 5 | $0.00030 | $0.00255 |
| Haiku 4.5 | $0.00015 | $0.00128 |
Grade A, and why
checkpoint-handoff scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 9d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 81 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Checkpoint Handoff
The failure mode this fixes
Long projects outlive context windows. The session that planned the work is never the session that finishes it — compaction, restarts, and tomorrow all guarantee that. And when the boundary hits without preparation, the next session starts by reconstructing instead of working: re-deriving the plan, re-discovering which tasks were done, re-learning the constraint the user stated forcefully forty turns ago, and — worst — confidently redoing or undoing finished work because nothing recorded that it was finished.
Conversation memory is the wrong place for project state. Summaries compress; the exact rule ("never X, always Y"), the precise task ordering, and the "we already tried that, it fails because Z" knowledge are exactly what compression loses. The fix is mechanical: before the boundary, the state gets written to durable files, structured for a cold reader.
When to Use This Skill
- The user says they're about to compact, or you notice the context getting heavy mid-project
- A session is ending with the project unfinished (i.e., almost every session of a real project)
- Work is being handed to another agent, a fresh session, or a parallel worker
- A phased plan is mid-flight — some tasks done, some in progress, some queued
- You were just asked "where were we?" — the skill fired one session too late
Don't use when: the task is single-session and will finish here; a checkpoint of a finished task is just a report. Don't checkpoint into the conversation itself — a summary message that will be compacted along with everything else is not a checkpoint.
Instructions
1. Write to durable files, not to the chat
The checkpoint lives where compaction can't touch it: a working doc in the repo (PLAN.md, a project doc, whatever the project already uses), agent memory if available, a tracked task list. If a working doc already exists, update it — a trail of stale plan files is its own failure mode.
2. Record state as facts, not narrative
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 9d ago First seen · 81 lines · 151 tokens per session scan A 788d28f08b01
checkpoint-handoff is a skill published in the GitHub repository mikestangdevs/craft-skills (4 stars, last pushed 3mo ago), licensed MIT. It adds 151 tokens to every session and 1,277 once invoked, about $0.0008 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
learn-from-fix
Capture Elixir/Ecto/LiveView lessons and Hex API rules. Use after corrections or when asked to document learning, record a lesson, prevent a fixed mistake, or remember package guidance with --library.
assigns-audit
Inspect LiveView socket assigns for memory bloat — missing temporaryassigns, unused assigns, unbounded lists needing streams, memory estimates. Use when LiveView memory grows or you need to add temporaryassigns.
recall
Recall prior work from past sessions — how a bug was fixed, what was decided, where a pattern lives. Use when asked 'have we done this before' or 'how did I fix X' in Elixir/Phoenix work. ccrider MCP when available, else git + solution docs.
compound-docs
Searchable Elixir/Phoenix/Ecto solution documentation system with YAML frontmatter. Builds institutional knowledge from solved problems. Use when consulting past solutions before investigating new issues.
context-anchoring
Manage per-feature living documents that capture decisions, constraints, and reasoning across AI sessions during active development. Scoped to feature-level work — design, implementation, bugfix, refactor — not for codebase-wide assessments or product-wide specifications (those define their own document lifecycles).…
learning-harvest
Manage the operational learnings lifecycle — load prior learnings to inform current work, harvest new patterns worth preserving, and keep the document tight over time. Provides a protocol for accumulating actionable patterns from practice that complement standards and defaults. Use when a workflow session completes…