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 arturseo-geo/claude-code-skills --skill context-engineeringgit clone --depth 1 https://github.com/arturseo-geo/claude-code-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/arturseo-geo/claude-code-skills/context-engineering)<a href="https://agentmods.dev/skills/arturseo-geo/claude-code-skills/context-engineering"><img src="https://agentmods.dev/badge/skills/arturseo-geo/claude-code-skills/context-engineering.svg" alt="Measured on agentmods" 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 | $0.00150 | $0.02468 |
| Opus 5 | $0.00075 | $0.01234 |
| Sonnet 5 | $0.00030 | $0.00494 |
| Haiku 4.5 | $0.00015 | $0.00247 |
Grade A, and why
context-engineering 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 8d 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 — 280 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Context Engineering Skill
How Claude Code's Context Window Works
Claude Code's 200k context window is split across fixed overhead and your conversation:
| Slot | Typical size | Notes |
|---|---|---|
| System prompt | ~2.7k tokens | Fixed |
| System tools | ~16.8k tokens | Fixed — all built-in tools |
| Custom agents | ~1.3k tokens | Per agent file loaded |
| Memory files (CLAUDE.md etc.) | ~7.4k tokens | Grows with your config |
| Skills (metadata) | ~0.5–1k tokens | ~100 tokens per skill description |
| Messages (conversation) | Grows → triggers compaction | Your actual work |
| Autocompact buffer | 33k tokens | Hardcoded — cannot be changed |
Working limit: compaction fires at ~167k tokens (200k minus 33k buffer). When it fires, your precise early context (variable names, exact errors, decisions) gets lossy-summarised.
Autocompaction Control
# Default: fires at ~83.5% capacity
# Override to trigger earlier (more frequent, less data loss per event)
export CLAUDE_AUTOCOMPACT_PCT_OVERRIDE=70
# Or later (longer sessions, more data lost when it fires)
export CLAUDE_AUTOCOMPACT_PCT_OVERRIDE=90
Set in .claude/settings.json to persist:
{
"env": {
"CLAUDE_AUTOCOMPACT_PCT_OVERRIDE": "70"
}
}
Strategic compaction: run /compact manually at logical checkpoints
(end of a feature, after a big refactor) instead of waiting for auto-fire.
This gives you control over what gets summarised vs lost.
Focused compaction: guide what gets preserved:
/compact focus on the auth refactor decisions and the API contract
For the full autocompaction deep dive (triggers, thresholds, anti-patterns), read references/autocompaction.md.
Reducing Static Context Overhead
CLAUDE.md diet
Every line in CLAUDE.md loads every session. Keep it under 100 lines and under 500 tokens.
The Router Pattern — turn CLAUDE.md into a routing table, not a manual:
## Project: myapp
Stack: Next.js 15, Prisma, PostgreSQL
Use pnpm. API routes in src/app/api/.
| Keywords | Skill |
|---|---|
| deploy, release, CI | deploy |
| test, coverage | testing |
| schema, migration | database |
What ships with it
4 files 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.
- 8d ago First seen · 280 lines · 150 tokens per session scan A 938fa6582554
context-engineering is a skill published in the GitHub repository arturseo-geo/claude-code-skills (11 stars, last pushed 5mo ago), licensed MIT. It adds 150 tokens to every session and 2,468 once invoked, about $0.0007 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 2026-08-30.
Other skills, from other repositories
continuous-learning-v2
Instinct-based learning system that observes sessions via hooks, creates atomic instincts with confidence scoring, and evolves them into skills/commands/agents.
memory-hygiene
Keep agent-written memory files from rotting. Audits /.claude/projects/ /memory/ for facts the tree now contradicts — stale version pins, passed sequence facts, vanished paths — plus MEMORY.md index drift, and enforces the memory format contract at write time. Use when the user says "audit my memory", "memory is…
learn-on-failure
Save a learning to project memory. Invoke automatically (without the user asking) whenever a task required more than one fix cycle to resolve — e.g. a test failed and needed a second attempt, a compile error required a correction, an API behaved unexpectedly, or an assumption proved wrong mid-task. Also invoke when…
letter-hooker
Erweitert automation-self-care um Letter Hooks, Preflight-Bootloader, Dokument-Traversierungsregeln und selbstheilende Prompt-Kontext- Anreicherung für KI-Agenten und CLIs ohne native, ereignisgetriebene JSON-Lifecycle-Hooks (wie Antigravity / Gemini CLI). Nutzen, wenn ein Agent Preflight-Regeln injizieren, vor…
hive.note-taking
Maintain a free-form scratchpad of decisions, extracted values, and open questions so context pruning doesn't lose anything you still need.
security-compliance
Guides security professionals in implementing defense-in-depth security architectures, achieving compliance with industry frameworks (SOC2, ISO27001, GDPR, HIPAA), conducting threat modeling and risk assessments, managing security operations and incident response, and embedding security throughout the SDLC.