mcpField
P
principle-guard-the-context-window

principle-guard-the-context-window

github.com/mdsmithaustinAgent workflows8159 installsUpdated 1d ago
Repository

About principle-guard-the-context-window

A skill for agents and developers that treats session context as a limited resource. It lays out ways to reduce unnecessary context use, delegate large inputs, and keep work scoped.

Key features

  1. 01Delegate large payloads for summaries
  2. 02Keep recurring references inline
  3. 03Limit work by phase and budget
  4. 04Request concise subagent reports
  5. 05Trim long file and command output
  6. 06Read only relevant playbook steps

Use cases

  1. 01Handling large documents or verbose tool output
  2. 02Planning multi-phase agent work
  3. 03Passing subagent findings to a lead
  4. 04Inspecting lengthy playbooks efficiently

Install

Copy the command and run it in a terminal.

$npx skills add https://github.com/mdsmithaustin/pstack --skill principle-guard-the-context-window

After installing, type/principle-guard-the-context-windowin Claude Code to confirm it loaded.

Bundled filesagents/openai.yaml

Details

From the project README

Guard the Context Window

The context window is finite and non-renewable within a session. Every token should be worth its cost.

Why: Context overflow degrades reasoning quality, creates compression artifacts, and halts progress.

Pattern:

  • Isolate large payloads. Route verbose outputs, screenshots, and large documents to subagents. The main context gets summaries, not raw data. A read-only scan or parse that returns a reduced finding runs on the routine role, unless the calling skill names another role. Use the pstack-harness skill's Spawn a role section to resolve the role.
  • Keep frequently used content inline. Templates and references used on every invocation belong in the skill file, not in separate files that cost a read each time.
  • Size phases and cap scope. Limit files per phase, set turn budgets, account for mechanism costs.
  • Return thin. When a subagent's report feeds your next step, its brief names where the full report goes, a file the subagent writes or the PR, and asks for only the verdict or status line, any blockers, and that path or URL. Read the full report only to act on a blocker. A subagent whose output is the deliverable you relay, such as a how explanation, a why synthesis, the trail reviewer's flags, or an arena cross-judge's scores, returns it in full.
  • Budget the lead's reads. Send a file or command output longer than about 150 lines to a subagent unless you will edit it. Cut command output with head, tail, grep, or --jq before it reaches you. Do not reread a file a how pass already mapped unless you are about to edit it. A read that a playbook or skill step assigns to you, such as Code review's changed files, Eval's candidate outputs, or show-me-your-work's transcript audit, is exempt.
  • Read steps, not whole playbooks. To open a worklist, print the playbook's prose before step 1 and its numbered steps with their indented sub-items (awk '/^[0-9]+\. /{p=1} !p && !/^#/ && NF {print; next} p && /^[^ ]/ && !/^[0-9]+\. /{exit} p' <playbook>) instead of the whole file. The prose before step 1 is the worklist's first item. When that prose says to read the playbook in full, read it in full.

Comments

Markdown supported