Agent Skill

Context Engineering skill

context-engineeringmvschwarz/openrigpackages/daemon/specs/agents/shared/skills/process/context-engineering/SKILL.md

Use when designing, reviewing, or debugging how an agent's context window gets filled, pruned, or shared — choosing what loads at boot versus on demand, sizing an install or an always-loaded file, fixing an agent that drifts, repeats itself, or forgets constraints mid-task, planning compaction or summarization, deciding single-agent versus subagents, engineering handoffs between agents, or picking a tool loadout. NOT for rewording a prompt's tone, choosing which model to pin, or debugging business logic — those are adjacent moments this skill does not serve. Historical 2024–2025 snapshot; not normative for present-day frontier models — see the Status section.

8.1kApache-2.0

always on

142 tokname + description

once used

7,182 tokSKILL.md body

bundled files

repo stars

8.1kGitHub

Install the Context Engineering skill

Skills CLI (Claude Code, Codex, Cursor and others)

npx skills add https://github.com/mvschwarz/openrig --skill context-engineering

Claude Code, by hand (all projects)

git clone --depth 1 https://github.com/mvschwarz/openrig.git /tmp/openrig
mkdir -p ~/.claude/skills
cp -R /tmp/openrig/packages/daemon/specs/agents/shared/skills/process/context-engineering ~/.claude/skills/context-engineering

For one project only, copy the folder into .claude/skills/ in that project instead. The agent picks it up when a task matches its description.

What it costs your context

Agent Skills load in three steps. Tokens counted with the o200k tokenizer; your model's count will differ a little.

When	What loads	Cost
Every session	Name and description	142 tok
When the agent uses it	The SKILL.md instructions	7,182 tok
Only if opened	No bundled files	—

Until it is used, context-engineering costs about 9.6× less than the median MCP server we measured (1,362 tokens on every turn). Installing all 94 skills from mvschwarz/openrig would load 6,757 tokens of descriptions. Skills vs MCP study

Checks

The instructions are over 5,000 tokens, the Agent Skills specification's guideline for SKILL.md.

Checked 2026-10-22 · method

What's inside

The headings of its SKILL.md. Read the full file on GitHub.

Traditional Context Engineering — 2024–2025 Snapshot
Status: provisional historical research snapshot
1. The mental model: what context engineering is
2. Core principles (and the why behind each)
3. The technique catalog
3.1 Progressive disclosure
3.2 Just-in-time retrieval (vs. pre-computed context)
3.3 Compaction and summarization
3.4 Context editing / tool-result clearing
3.5 Structured note-taking (agentic memory)
3.6 Sub-agents and context isolation
3.7 Structured handoffs and delegation contracts
3.8 Tool loadout and tool design
3.9 Context budgets and caching discipline
4. Failure modes
5. How the best teams operate
6. Decision guide
7. Contested territory and thin spots (curator flags)
Sources

More skills from mvschwarz/openrig

All skills in this repo

agent-browsermvschwarz/openrig
Browser automation CLI for AI agents. Use when the user needs to interact with websites, including navigating pages, filling forms, clicking buttons, taking screenshots, extracting data, testing web apps, or automating any browser task. Triggers include requests to "open a website", "fill out a form", "click a button", "take a screenshot", "scrape data from a page", "test this web app", "login to a site", "automate browser actions", or any task requiring programmatic web interaction.
8.1k108 tok idle · 3,477 in use
agent-operated-softwaremvschwarz/openrig
Use when designing, building, operating, or diagnosing an ongoing application whose live backend or control loop includes OpenRig agents, including applications with a Markdown, YAML, or JSON agent control plane or a thin surface over specialist agent roles.
8.1k51 tok idle · 910 in use
agent-operated-workflowsmvschwarz/openrig
Use when a bounded real-world procedure has a deterministic happy path but brownfield, variable, or partially knowable state; when an operation must resume from verified evidence; or when deciding whether agent judgment or ordinary code should own a procedure's control loop.
8.1k55 tok idle · 730 in use
agent-startersmvschwarz/openrig
Use when creating, refreshing, packaging, inspecting, promoting, or deprecating a named per-seat starting point — Agent Starter manifest authoring, the 6-state lifecycle (captured → named → inspectable → used → promoted → deprecated), provenance honesty, and refusal rules. NOT a VM image; a managed starting point composed from agent role + startup context + optional native session source + provenance.
8.1k84 tok idle · 1,317 in use
agent-startup-and-context-ingestionmvschwarz/openrig
Use when designing or auditing how an agent becomes useful after launch — AGENTS.md overlays, role files, skills, rig specs, workflow specs, startup checklists, refocus messages, "rig context" surface. Covers the 4 failure modes that make startup context fail (old rig spec misses current operating mode; current agents never told about new guidance; startup file as dumping ground; orchestrator transmits implementation without preserving product intent).
8.1k95 tok idle · 1,276 in use
applying-a-permission-policymvschwarz/openrig
Use before launching a team during agent-guided setup, or when a user asks to configure OpenRig command permissions, reduce repeated native approval prompts, or apply a selected rig/seat permission policy.
8.1k46 tok idle · 2,819 in use
backlog-capturemvschwarz/openrig
Quickly capture product ideas, feature requests, or insights from meetings and conversations. Rapid documentation with smart categorization and deduplication.
8.1k31 tok idle · 434 in use
claude-compaction-restoremvschwarz/openrig
Use when a Claude Code session is about to compact, has just compacted, reached its context limit, resumed after /compact, or must rebuild its working picture of a long-running seat from its own restore map, its JSONL transcript and the files it touched.
8.1k260 tok idle · 3,982 in use