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 mattbaconz/signal --skill signal-ckptgit clone --depth 1 https://github.com/mattbaconz/signalWrote 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/mattbaconz/signal/signal-ckpt)<a href="https://agentmods.dev/skills/mattbaconz/signal/signal-ckpt"><img src="https://agentmods.dev/badge/skills/mattbaconz/signal/signal-ckpt/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/mattbaconz/signal/signal-ckpt"><img src="https://agentmods.dev/badge/skills/mattbaconz/signal/signal-ckpt.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 | $0.00080 | $0.01402 |
| Opus 5.5 | $0.00032 | $0.00561 |
| Sonnet 5.5 | $0.00016 | $0.00280 |
| Haiku 4.5 | $0.00008 | $0.00140 |
Grade A, and why
signal-ckpt 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 2mo 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 — 185 lines — stays where its author put it; the contents beside it link to each section on GitHub.
⚡ signal-ckpt — Manual Checkpoint
Collapse the session into a ≤50 token state atom. Drop all prior history. Resume from the atom.
Invocation Triggers
Activate when user says any of:
/signal-ckpt"checkpoint","compress context","summarize session""save state","compress history","ckpt"- Auto-fires in SIGNAL-3 mode every 5 turns
Output Format
CKPT[N]:
§project={name} §stack={tech}
progress=[{task}✓, {task}✗, {task}∅, {task}/]
blockers=[{issue}⊥{component}]
next={next_task}
§decisions=[{decision}✓]
N = checkpoint sequence number, starting at 1. Increment on each checkpoint.
Minimal valid checkpoint (when session is simple):
CKPT[1]:
§project=my-app
progress=[auth✓, api/]
next=finish api endpoints
Full checkpoint (active blockers, decisions, aliases):
CKPT[3]:
§project=data-pipeline §stack=python+airflow
[X1]=schema_validation_step [X2]=postgres_loader
progress=[ingest✓, transform✓, X1✗, X2∅]
blockers=[X1⊥schema_mismatch]
next=fix schema_mismatch in X1
§decisions=[use_upsert✓]
Hard Rules
-
≤50 tokens total. Count ruthlessly. If over, compress further: merge related tasks, abbreviate names, drop resolved items.
-
Checkpoint replaces conversation history. After this fires, all prior turns can be dropped. The checkpoint IS the history.
-
Only active blockers. Resolved blockers are gone. They do not appear anywhere in the checkpoint.
-
Only future-relevant decisions.
§decisionsonly includes decisions that affect work not yet complete. Past decisions with no forward implication are dropped. -
nextis singular. One task. The most immediate pending action. Not a backlog. -
Carry active aliases forward. Any
[Xn]aliases still in use must appear in the checkpoint. Aliases for resolved items are dropped.
Collapse Algorithm
When generating a checkpoint:
- Identify all tasks mentioned in the session. Assign each a status from its last known state (
✓✗∅/⊥).
What ships with it
1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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.
- 2mo ago First seen · 185 lines · 80 tokens per session scan A f2c8d8099f1e
signal-ckpt is a skill published in the GitHub repository mattbaconz/signal (11 stars, last pushed 4mo ago), licensed MIT. It adds 80 tokens to every session and 1,402 once invoked, about $0.0003 per session on Opus 5.5. 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.
signal-ckpt compared with skills from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| signal-ckptmattbaconz/signal | Skill | 11 | 4mo ago | MIT | A | 80 | Claude Code |
| lov-auto-contextlovstudio/skills | Skill | 72 | today | MIT | A | 112 | Claude Code |
| lov-ataru-indexinglovstudio/skills | Skill | 72 | today | MIT | A | 55 | Claude Code |
| lov-search-chatlovstudio/skills | Skill | 72 | today | MIT | A | 53 | Claude Code |
| lov-memory-addlovstudio/skills | Skill | 72 | today | MIT | A | 43 | Claude Code |
| lov-memory-searchlovstudio/skills | Skill | 72 | today | MIT | A | 42 | Claude Code |
| learncraigcossairt/trellis | Skill | 5 | yesterday | MIT | A | 77 | Claude Code, Codex |