Context Engineering Advisor skill: what it does and how to install it
Diagnoses bloated AI agent workflows and recommends context boundaries, retrieval and memory practices to improve reliability, focus and cost.
Summary generated from the skill's documentation.
Install
$ npx skills add deanpeters/Product-Manager-Skills --skill context-engineering-advisorRun it in a terminal. If your agent is already running, start a new session so it picks the skill up.
About this skill
What it does. Guides product managers through an interactive assessment of unreliable AI workflows, using five diagnostic questions to separate context stuffing from deliberate context engineering. It recommends what to persist or retrieve, who should own boundaries, structural fixes, memory layers, and a Research -> Plan -> Reset -> Implement cycle.
When to use it. Use it when large context dumps, retries, vague outputs or rising token costs suggest poor information architecture. It is not for choosing AI tools, writing prompts, or fixing workflows that already work well.
History
Repo stars
7.7k+512 since 5 Oct 2026
Stars are counted for the whole repository, which holds 19 skills.
Show as a table
| Date | Repo stars |
|---|---|
| 26 Oct 2026 | 7,678 |
| 25 Oct 2026 | 7,662 |
| 24 Oct 2026 | 7,637 |
| 23 Oct 2026 | 7,617 |
| 22 Oct 2026 | 7,589 |
| 21 Oct 2026 | 7,570 |
| 20 Oct 2026 | 7,542 |
| 19 Oct 2026 | 7,503 |
| 18 Oct 2026 | 7,461 |
| 17 Oct 2026 | 7,441 |
| 16 Oct 2026 | 7,421 |
| 15 Oct 2026 | 7,398 |
| 14 Oct 2026 | 7,363 |
| 13 Oct 2026 | 7,328 |
| 12 Oct 2026 | 7,278 |
| 11 Oct 2026 | 7,238 |
| 10 Oct 2026 | 7,224 |
| 9 Oct 2026 | 7,214 |
| 8 Oct 2026 | 7,198 |
| 7 Oct 2026 | 7,186 |
| 6 Oct 2026 | 7,176 |
| 5 Oct 2026 | 7,166 |
Installs
2.7k
Tracking since . A chart appears once there are 7 days of data.
Installs via skills.sh
Similar skills
- Caveman LearnReviews Caveman token-cost reports and applies consent-gated edits to reduce repeated context and heavy configuration.
- Suggestion Box (Warp)Reports concrete, generalisable friction in agent tooling and workflows to an internal suggestion box with structured, safe context.
- Caveman CompressCompresses a memory file such as CLAUDE.md into caveman format to save input tokens, keeping a readable backup.