checkpoint

Instructions for saving the current state of a long coding task at a clean stopping point.

In plain words
What is it for?
Use it after a completed work unit, before a session ends, or when the user asks to save progress.
Why use it?
It preserves what was done, what remains, and relevant project knowledge so work can continue after a break or context limit.

Skill for Claude CodeCodex

Part of the containuity plugin — 1 skill, 2 hooks shipped together

Install

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.

agentmods
npx agentmods add skills/vonargo/containuity/checkpoint
Any agent
npx skills add vonargo/containuity --skill checkpoint
Clone the repo
git clone --depth 1 https://github.com/vonargo/containuity

Made for: Claude Code, Codex.

Or install containuity, the plugin that ships this one along with the rest of its 1 skill, 2 hooks.

Per session 65 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,969 The whole file, excluding the scripts and references it only reads on demand.
Security scan B 1 finding. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5 $0.00065 $0.01969
Opus 5 $0.00032 $0.00984
Sonnet 5 $0.00013 $0.00394
Haiku 4.5 $0.00006 $0.00197

Measured 2d ago against content hash 496747049623, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade B, and why

checkpoint scanned grade B with 1 finding 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 2d 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.

Reads agent configuration directoriesmediumAgent snooping

.claude/, .codex/, .gemini/ hold keys, settings and other credentials a mod has no legitimate need for.

!`cat .checkpoint.config 2>/dev/null || cat .claude/checkpoint.config 2>/dev/null || echo "NO_CONFIG"`
skills/checkpoint/SKILL.md · 112 lines

How it starts

The opening of the file, as written. The whole thing — 112 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Checkpoint

Run this checklist at a clean breakpoint to hold continuity across sessions and compactions. Perform each step — do not just print the checklist. Create one todo per step (1–4) and work them in order.

Seat configuration (auto-loaded)

!`cat .checkpoint.config 2>/dev/null || cat .claude/checkpoint.config 2>/dev/null || echo "NO_CONFIG"`

Resolve these keys from the block above (fall back to the defaults, and tell the user which you used):

Key Default Meaning
LIVING_DOC docs/DESIGN.md The source-verified "how" doc you keep non-stale.
MEMORY_DIR .claude/memory Where durable memories live (one fact per file + MEMORY.md index).
COMMIT_MODE offer offer = propose the commit and wait for OK. auto = commit without asking.
COMMIT_CONVENTION infer infer from git log, or a freeform note overriding it.

If NO_CONFIG: use the defaults, name them to the user, and mention they can run the worked example in the plugin README to create .checkpoint.config once per seat.

Current state (auto-collected)

Outstanding work:

!`git status --short 2>/dev/null || echo "not a git repo"`

Recent commits — infer author, message style, and co-author policy from these (do not impose your own):

!`git log -8 --pretty='format:%an <%ae> | %s' 2>/dev/null || echo "no git history"`

The checklist

1. Commit outstanding work

  • If there is nothing to commit, say so and skip.
  • Match the repo's existing convention inferred from the log above: author, message style (imperative vs. scoped vs. conventional-commits), and co-author trailer policy. If the log shows no co-author trailers, do not add one; if it shows them, match the form.
  • COMMIT_MODE=offer (default): stage the relevant changes, show the exact message you propose, and wait for approval. Do not commit until the user authorizes commits for this repo (in this session or via COMMIT_MODE=auto).
  • COMMIT_MODE=auto: commit directly using the inferred convention.
  • Never git push unless the user explicitly asks.

Read the full file on GitHub · 112 lines

Changes

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.

  1. 2d ago First seen · 112 lines · 65 tokens per session scan B 496747049623

Subscribe to this mod's changes

checkpoint is a skill published in the GitHub repository vonargo/containuity (1 stars, last pushed 2mo ago), licensed MIT. It adds 65 tokens to every session and 1,969 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it B with 1 finding (reads agent configuration directories). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.

Related

Other skills, from other repositories

civitai-mcp-ultimate

Ultimate MCP server for Civitai — search models, browse top images with prompts, download LoRAs/Checkpoints, analyze trends. Use when user asks about AI models, LoRAs, checkpoints, Stable Diffusion, Flux, image generation prompts, or Civitai.

timoncool/civitai-mcp-ultimate · 65 tokens

container-manager-podman-operations

Rootless Podman pod/kube operations via the container-manager-mcp MCP server — pods (create/list/stats/top/inspect/logs/stop/rm), Kubernetes YAML interop (generate/play kube), checkpoint/restore, pod-scoped networks and volumes, health checks, and system prune. Use when the agent must drive Podman pod-level workloads…

Knuckles-Team/container-manager-mcp · 114 tokens

Training

Guides fine-tuning and post-training work with explicit data, objective, hardware, and rollback assumptions.

agentic-in/elephant-agent · 23 tokens

agent-rollback

Use the agent-rollback MCP server or ar CLI to snapshot the workspace, list/show/diff checkpoints, and roll back unwanted changes. Use this skill whenever the user mentions checkpoints, rollback, undo, revert, restore, "go back to", "what changed", "I broke something", "this isn't working", or asks to inspect or…

Nainish-Rai/agent-rollback · 139 tokens

context-bridge

Cross-session task memory and stagnation detection for Claude Code, backed by a local context-bridge server that auto-starts on session start. Use when hook output tagged [context-bridge] appears, when the user says "checkpoint", "sync context", or "remember this for next session", when a task keeps reappearing…

pushkalkumar/context-bridge · 89 tokens

repo-checkpoint

Cross-agent repo-local progress checkpoint workflow for any git repository. Use this skill whenever the user says save progress, checkpoint, snapshot the current lane, capture key chat context, write down what matters before closing, or wants the next session to resume immediately from Codex, Claude Code, opencode, or…

hotalexnet/agent-checkpoint · 68 tokens