Skills DirectorySkills Directory
SkillsLearnSecurityCategoriesDocsBlogPro
Sign InSubmit Skill
Skills Directory

Security-tested agent skills for Claude, coding agents, and AI workflows.

Directory

  • Browse Skills
  • All Skills A–Z
  • Claude Skills
  • Claude Code Skills
  • Agent Skills
  • Categories
  • Authors
  • Submit a Skill

Learn

  • Learn Hub
  • Install Claude Skills
  • Write SKILL.md
  • Skills vs MCP
  • Directories Compared

Security

  • Security
  • Methodology
  • Secure Claude Skills
  • Security Badges
  • Chrome Extension
  • Skill Manager

Company

  • About
  • Community
  • Blog
  • API Docs
  • Advertise

2026 Skills Directory. All rights reserved.

ProTermsPrivacyRefunds
Back to skills

Context Mode

ASecurity

Context-mode is an MCP server plugin that reduces context window usage by ~98% through sandboxed tool routing. MUST be applied whenever Claude is running inside a context-mode-enabled session — it is not optional. Trigger when: context-mode MCP server is present in the session; running Bash commands that produce more than 20 lines of output; reading files for analysis; fetching web pages and searching them; batch execution of multiple queries; or when you need to prevent raw tool outputs from...

3 stars
0 votes
0 copies
0 views
Added 9/29/2026
ai-agentsbashdebugginggit

Works with

mcp

Security Analysis

A100/100

Scanned 9/29/2026

$npx -y skills add Tekkiiiii/the-agency --skill context-mode --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Context Mode?

Add the live security badge to your README — it updates automatically with every re-scan.

Security grade badge for Context Mode
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/tekkiiiii-context-mode/badge)](https://www.skillsdirectory.com/skills/tekkiiiii-context-mode)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
Files
SKILL.md
---
name: context-mode
description: |
  Context-mode is an MCP server plugin that reduces context window usage by ~98% through
  sandboxed tool routing. MUST be applied whenever Claude is running inside a context-mode-enabled
  session — it is not optional. Trigger when: context-mode MCP server is present in the session;
  running Bash commands that produce more than 20 lines of output; reading files for analysis; fetching
  web pages and searching them; batch execution of multiple queries; or when you need to prevent raw
  tool outputs from flooding the context window. Key capabilities: routes Bash to ctx_execute /
  ctx_batch_execute (captures output, prevents bloat); routes Read to ctx_execute_file (raw content
  never enters context); routes WebFetch to ctx_fetch_and_index then ctx_search (fetch once, query
  indexed content on demand); enforces strict output rules (under 500 words, write artifacts to files,
  return only path + 1-line description). Ideal for: working in long-running sessions with large
  codebases; running multi-step debugging workflows that produce verbose output; any task where
  context window preservation is critical. Also for: batch operations over large project directories;
  running analysis across many files without consuming context space.
---

# context-mode

Context-mode routes data-heavy tool calls through sandboxed MCP tools, preventing raw outputs from flooding the context window.
Applies automatically when context-mode MCP server is running in the session. Always enforced — do not skip.

---

## When Invoked
- Always active when context-mode MCP server is present in the session
- Detected automatically by the session environment
- No explicit user trigger needed — enforce at all times

---

## Tool Routing Rules

Route every tool call to its sandbox equivalent:

| Instead of | Use | Why |
|---|---|---|
| Bash commands (>20 lines output) | `batch_execute` or `ctx_execute` | Captures output, prevents context bloat |
| Read for analysis | `ctx_execute_file` | Raw content never enters context |
| WebFetch | `ctx_fetch_and_index` then `search` | Fetch, index, then query by search |
| Follow-up searches | `ctx_search` | Query indexed content on demand |
| Batch commands/queries | `ctx_batch_execute` | One call replaces many steps |

---

## Forbidden Patterns

- **NEVER use Bash** for commands producing >20 lines of output — use `ctx_execute` or `ctx_batch_execute`
- **NEVER use Read** for analysis tasks — use `ctx_execute_file`
- **NEVER use WebFetch** — use `ctx_fetch_and_index` then `ctx_search`
- **NEVER use curl/wget in Bash** — use `ctx_execute` or `ctx_fetch_and_index`
- Bash is ONLY for: git, mkdir, rm, mv, navigation, and short commands (<20 lines)

---

## Output Rules

- Keep responses under 500 words
- Write artifacts (code, configs) to files — never return them as inline text
- Return only: file path + 1-line description

---

## Local Dev Notes (for working on context-mode itself)

- Hook dispatcher: `context-mode hook <platform> <event>`
- Delete `server.bundle.mjs` during local dev to load changes from `build/server.js`
- Version must sync across 4 files
- Tests: `npm test` (Vitest, parallel), `npm run test:watch`
- TDD workflow: failing test first → minimum code → refactor
- Add tests to existing test files covering the same domain — do not create new test files

Attribution

TekkiiiiiTekkiiiii
View sourceSee grades on GitHubMore from Tekkiiiii →
SSkills DirectorySkills Directory

Ship a skill? Prove it's safe.

Free 120-pattern security scan, letter grade, and an embeddable README badge.

Submit a skill

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments (0)

No comments yet. Be the first to comment!

SSkills DirectorySkills Directory

Ship a skill? Prove it's safe.

Free 120-pattern security scan, letter grade, and an embeddable README badge.

Submit a skill

Related Skills

Caveman

Terse caveman voice: answer first, fluff gone, every technical fact kept. Use for /caveman, "caveman mode", "talk like caveman", "be brief", "less tokens". Stays on until "stop caveman" or "normal mode".

1100021 votes

Hyperplan

Adversarial multi-agent planning skill. Self-orchestrates 5 hostile category members (unspecified-low, unspecified-high, deep, ultrabrain, artistry) via team-mode for ruthless cross-critique debate, distills only the defensible insights, then MANDATORILY hands the distilled insight bundle to the `plan` agent for executable plan formalization. Use when planning needs maximum rigor and surfacing of weak assumptions, blind spots, and over-engineering. Triggers: 'hyperplan', 'hpp', '/hyperplan', ...

698431 votes

Writing Skills

Create and manage Claude Code skills in HASH repository following Anthropic best practices. Use when creating new skills, modifying skill-rules.json, understanding trigger patterns, working with hooks, debugging skill activation, or implementing progressive disclosure. Covers skill structure, YAML frontmatter, trigger types (keywords, intent patterns), UserPromptSubmit hook, and the 500-line rule. Includes validation and debugging with SKILL_DEBUG. Examples include rust-error-stack, cargo-dep...

3931 votes

Mcp Code Execution

Routes multi-tool workflows through MCP servers for large datasets and pipelines. Use when Bash tool overhead is limiting throughput on data-heavy tasks.

3421 votes

catchup

Recovers the conversation and failed tool calls of a previous Codex, Amp, Claude Code, Antigravity, Cline, Copilot CLI, Cursor, DeepSeek Harness, Grok Build, Kimi, OpenCode, Pi Agent, or ZCode session. Use when the user says "catch up", "what did the last session do", "get me up to speed", "I switched agents", asks to recover/summarize a previous session before continuing, or asks to diagnose or report a catchup failure. Do NOT use for the current conversation, git history, or any non-agent log.

741 votes
View all in ai-agents →