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 mlopscommunity/Coding-Agents-Conference-skills --skill context-window-managementgit clone --depth 1 https://github.com/mlopscommunity/Coding-Agents-Conference-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/mlopscommunity/coding-agents-conference-skills/context-window-management)<a href="https://agentmods.dev/skills/mlopscommunity/coding-agents-conference-skills/context-window-management"><img src="https://agentmods.dev/badge/skills/mlopscommunity/coding-agents-conference-skills/context-window-management.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 | $0.00037 | $0.01620 |
| Opus 5 | $0.00018 | $0.00810 |
| Sonnet 5 | $0.00007 | $0.00324 |
| Haiku 4.5 | $0.00004 | $0.00162 |
Grade A, and why
context-window-management 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 — 130 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Context Window Management
Overview
Techniques for managing what goes into the context window so the model stays focused and produces high-quality output. The core insight is counterintuitive: less context produces better results. Stuffing the context window with everything available degrades performance. Instead, save state to files, keep prompts short, and restart sessions when confused.
Core principle: The context window is expensive working memory, not cheap storage. Offload state to files and load only what you need right now.
When to Use
- At the start of any multi-step implementation (plan your context budget before you begin)
- When agent output quality drops mid-conversation (hallucinations, repetition, ignoring instructions)
- When combining multiple prompt sources (system prompt, CLAUDE.md, tool results, user instructions)
- When tool calls return large payloads (logs, file contents, API responses)
- When a task requires more than 3-4 back-and-forth exchanges
When NOT to Use
- Single-shot questions with short answers
- Tasks that genuinely need the full context loaded (e.g., cross-file refactors where all files fit comfortably)
- When you are already under 40% context utilization
Common Mistakes
| Mistake | Why it's wrong |
|---|---|
| Dumping entire file contents into context | Large tool results push out earlier instructions. Offload to a file and show only the first ~100 lines. Harrison Chase: "Load the first hundred lines, let the agent ask for more." |
| Writing 50+ instructions in a single prompt | The model starts dropping instructions past ~40. Dex: "Keep individual prompts under 40 instructions." Budget across all sources. |
| Using the system prompt for control flow logic | Prompts are for goals and constraints, not if/else branching. Use actual code for control flow. Dex: "Don't use prompts for control flow -- use actual control flow." |
| Continuing a confused conversation instead of restarting | Accumulated confusion compounds. Josh: "Clear and restart when confused. More context is not always better." |
| Treating the context window as a notebook | Important state belongs in static markdown files on disk, not carried in the conversation. Context is volatile; files are persistent. |
| Ignoring total instruction count across sources | Your system prompt, CLAUDE.md, tool descriptions, and user message all share one budget of ~150-200 instructions max. Going over means something gets silently dropped. |
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 · 130 lines · 37 tokens per session scan A f64b07afff60
context-window-management is a skill published in the GitHub repository mlopscommunity/Coding-Agents-Conference-skills (37 stars, last pushed 5mo ago), licensed Apache-2.0. It adds 37 tokens to every session and 1,620 once invoked, about $0.0002 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
media-ingest
Ingest video, audio, PDF, book, screenshot, and GitHub repo content into the brain. Multi-format handling with entity extraction and backlink propagation. Covers video-ingest, youtube-ingest, and book-ingest subtypes.
mem0-oss-to-platform
Plan and then execute a migration of a project from the mem0 open-source / self-hosted SDK (the local Memory class) to the mem0 Platform / hosted / managed SDK (the MemoryClient class). Use this whenever a developer wants to move, switch, or migrate their mem0 usage off OSS/self-hosted to the hosted API — e.g.…
Cortex
Operate Cortex, the LifeOS memory system — the typed Knowledge Archive (People, Companies, Ideas, Research with typed related: links) plus recall of prior work sessions, ISAs, and conversations. Search, add, harvest, develop, ingest, distill, graph-navigate, recall. USE WHEN cortex, knowledge, knowledge base, search…
memory
Use when the user asks to remember, recall, forget, update, search, or inspect durable OpenSquilla memory, including profile facts in USER.md and long-term notes in MEMORY.md or memory//.md.
ha-data-stores
Map of Hope Agent's local data stores and safe read-only query workflow. Use when the user asks where Hope Agent stores data, wants to inspect sessions/messages/memory/logs/background jobs/knowledge indexes/settings, asks the model to query local app data, or debugging requires checking persisted state. Trigger…
establishing-project-context
Use when the user asks to establish shared project language, or project work exposes a conflicting, renamed, or deprecated domain term that needs active semantic modeling. Routine small tasks stay on the fast path.