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 agentmods add commands/sliamh11/deus/checkpointgit clone --depth 1 https://github.com/sliamh11/DeusWhat 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 | $0.00000 | $0.00522 |
| Opus 5 | $0.00000 | $0.00261 |
| Sonnet 5 | $0.00000 | $0.00104 |
| Haiku 4.5 | $0.00000 | $0.00052 |
Grade A, and why
checkpoint 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 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.
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.
What it actually says
Save a mid-session checkpoint — for continuity between sessions of the same day.
First, resolve the vault path by reading ~/.config/deus/config.json and using the vault_path value. If the env var DEUS_VAULT_PATH is set, use that instead. All paths below use $VAULT to mean this resolved path.
Use before /compact when you want to resume exactly where you left off later that day. /resume will automatically load today's checkpoint.
Step 1 — Write checkpoint:
Identify: what decisions or intermediate conclusions have been reached in this session that are NOT yet saved in a session log?
Write to: $VAULT/Checkpoints/YYYY-MM-DD-HH.md (Use current date and 24h hour. Create the Checkpoints/ folder if it doesn't exist.)
Use exactly this format:
---
type: checkpoint
created: YYYY-MM-DDTHH:MM
session_topic: short-slug
decisions:
- "decision made so far (≤12 words)"
in_progress:
- "what we are actively working on right now"
next_action: "the exact next step to take after resuming"
context_refs:
- "file path or resource name needed to continue"
---
## Mid-Session State
3–5 sentences explaining where we are, what has been decided, and what comes next.
Write as if explaining to yourself after a 30-minute break.
Keep the checkpoint under 25 lines total. This is what /resume will load on the next session if it's the same day.
Step 2 — Confirm the checkpoint path was written.
Step 3 — Output context primer.
Before running /compact, output the following block verbatim (filling in values from current session state). This is a "compaction seed" — structured content near the end of conversation that the compaction algorithm will preserve as high-signal.
---BEGIN CONTEXT PRIMER---
## Active Task
[1 sentence: what we are working on right now]
## Session Decisions
[Bulleted list: decisions made in THIS session, max 5]
## Key Files
[Bulleted list: file paths actively being modified or referenced]
## Pending
[Bulleted list: what still needs to be done, max 3 items]
## Resume Hint
[1 sentence: if resuming after compaction, start by doing X]
---END CONTEXT PRIMER---
Step 4 — Tell the user: "Checkpoint saved. Run /compact now to compact the context."
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.
- 2d ago First seen · 63 lines · 0 tokens per session scan A 1be7a119f5cc
checkpoint is a command published in the GitHub repository sliamh11/Deus (51 stars, last pushed 8d ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 522 tokens. 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 2026-08-30.
Other commands, from other repositories
toon-decode
Convert TOON v2.0 format back to JSON using the Zig decoder.
toon-encode
Convert JSON to TOON v2.0 format using the Zig encoder.
toon-validate
Validate TOON v2.0 files for correctness.
memcord-save-progress
Auto-summarize and save conversation progress.
memcord-save
Save current conversation to memory.
memcord-close
Deactivate memory slot and end session.