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 kadiresen/context-bank --skill compactgit clone --depth 1 https://github.com/kadiresen/context-bankWrote 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/kadiresen/context-bank/compact)<a href="https://agentmods.dev/skills/kadiresen/context-bank/compact"><img src="https://agentmods.dev/badge/skills/kadiresen/context-bank/compact/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/kadiresen/context-bank/compact"><img src="https://agentmods.dev/badge/skills/kadiresen/context-bank/compact.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 | $0.00024 | $0.00170 |
| Opus 5.5 | $0.00010 | $0.00068 |
| Sonnet 5.5 | $0.00005 | $0.00034 |
| Haiku 4.5 | $0.00002 | $0.00017 |
Grade A, and why
compact scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured yesterday.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
What it actually says
- Run
npx -y [email protected] compact "$CLAUDE_PROJECT_DIR" --dry-runand show the user which files would change and what would be archived. - Ask for confirmation. Only after a clear yes, run
npx -y [email protected] compact "$CLAUDE_PROJECT_DIR" --yes. - Skim the new
.ai/active-context.md. If important current work was cut, restore those few lines by hand from the archive file it points to. - Summarize what moved. Do not commit; the user reviews the diff.
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.
- yesterday Changed 506b5f4bf539
- 2d ago Changed ab34fec453fc
- 14d ago First seen · 11 lines · 24 tokens per session scan A d35190f33669
compact is a skill published in the GitHub repository kadiresen/context-bank (9 stars, last pushed 2d ago), licensed MIT. It adds 24 tokens to every session and 170 once invoked, about $0.0001 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 2026-09-27.
Other skills, from other repositories
context
CONTEXT: Cognitive Order Normalized in Transformer EXtract Truncated. Cross-model context handoff via Progressive Density Layering, MLDoE expert compression, Japanese semantic density, and Negentropic Coherence Lattice validation. Creates portable carry-packets that transfer cognitive state between AI sessions. Use…
faf-context
Use FAF project context tools on this server — init, score, sync, discover; claim equals wire.
memseek-remember
Persist a user-confirmed fact, preference, constraint, or decision in Memseek project memory.
atomicmemory
Persistent semantic memory across Claude Code sessions — user preferences, project context, prior decisions, codebase facts. Call memorysearch before answering questions that reference past work. Call memoryingest after the user shares durable facts.
memseek-explain
Audit why Memseek recalled a claim by opening its evidence and replaying the original session when needed.
atomicmemory-cli
Use the installed AtomicMemory CLI for memory search, ingestion, packaging, diagnostics, and agent-safe JSON output.