context-engineering

github:addyosmani/agent-skills · first seen 2026-10-30 · installs 65.1K · source · badge

Optimizes agent context setup. Use when starting a new session, when agent output quality degrades, when switching between tasks, or when you need to configure rules files and context for a project.

Works with

ci-cd-and-automation
Shares configure, lint, quality, tsc
0.18
frontend-ui-engineering
Shares tsx, engineering, quality, button
0.17
figma-create-design-system-rules
Shares conventions, rules, cursor, project
0.16
planning-and-task-breakdown
Shares spec, too, tasks, entire
0.15
using-agent-skills
Shares engineering, starting, quality, development
0.15

Used with

ci-cd-and-automation
Shares configure, lint, quality, tsc
0.18
frontend-ui-engineering
Shares tsx, engineering, quality, button
0.17
planning-and-task-breakdown
Shares spec, too, tasks, entire
0.15
setup-pre-commit
Shares lint, configure, hooks, commit
0.14
web-artifacts-builder
Shares tailwind, vite, bundle, react
0.14

Skill Health 100 (freshness and quality) estimated by backskills

Listed, not yet scored. This skill has no public model test yet. It is in the scoring queue; submitted skills go first, then the most-installed.

Freshness 100 · Quality 100 · Efficacy not yet tested

Version history

2026-10-31T09:00:38+00:00 · sha256 5ee2929d3ade
2026-10-30T20:36:17 · sha256 5bf1e3806f20

Diff at 2026-10-31T09:00:38+00:00

--- previous
+++ current
@@ -199,7 +199,7 @@
 
 The context window is not a filing cabinet — it's a working desk. As a session runs, conversation history, tool output, and exploration accumulate. Most of it becomes deadweight. Budget proactively: waiting until the window is full causes abrupt quality drops; managing regularly keeps the agent coherent through long tasks.
 
-**Start trimming at 75% capacity, not 100%.** By the time the window is genuinely full, the model's attention is already fragmented across too many signals. The 75% threshold gives room to compress gracefully rather than cut desperately mid-task.
+**Compact well before the host's limit; the threshold varies by host.** By the time the window is genuinely full, the model's attention is already fragmented across too many signals. The margin gives room to compress gracefully rather than cut desperately mid-task.
 
 ### What to cut first
 
@@ -321,7 +321,7 @@
 | Missing examples | Agent invents a new style instead of following yours | Include one example of the pattern to follow |
 | Implicit knowledge | Agent doesn't know project-specific rules | Write it down in rules files — if it's not written, it doesn't exist |
 | Silent confusion | Agent guesses when it should ask | Surface ambiguity explicitly using the confusion management patterns above |
-| Context cliff | Waiting until the window is full before managing it — attention fragments and output quality drops abruptly at the limit | Start trimming at 75% capacity; compress rather than cut |
+| Context cliff | Waiting until the window is full before managing it — attention fragments and output quality drops abruptly at the limit | Compact well before the host's limit; compress rather than cut |
 
 ## Common Rationalizations