Context Engineering AdvisorSAFE

Product Management skills framework built on battle-tested methods for Claude Code, Cowork, Codex, and AI agents.

The instruction file, verbatim from the audited commit — this is the text the model reads, and the surface the audit's instruction layer examines. Quoted here so you can judge it without cloning anything.

---
name: context-engineering-advisor
argument-hint: "[AI workflow to diagnose]"
description: Diagnose context stuffing vs. context engineering. Use when an AI workflow feels bloated, brittle, or hard to steer reliably.
intent: >-
  Guide product managers through diagnosing whether they're doing **context stuffing** (jamming volume without intent) or **context engineering** (shaping structure for attention). Use this to identify context boundaries, fix "Context Hoarding Disorder," and implement tactical practices like bounded domains, episodic retrieval, and the Research→Plan→Reset→Implement cycle.
type: interactive
theme: ai-agents
best_for:
  - "Diagnosing context stuffing vs. context engineering in your AI workflows"
  - "Building better memory and retrieval architecture for AI agents"
  - "Improving AI output quality through structured context design"
scenarios:
  - "My AI outputs are mediocre even though I'm giving it lots of information — diagnose what's wrong"
  - "I want to architect context properly for a multi-step AI workflow in my product team"
estimated_time: "15-20 min"
---

## Purpose

Guide product managers through diagnosing whether they're doing **context stuffing** (jamming volume without intent) or **context engineering** (shaping structure for attention). Use this to identify context boundaries, fix "Context Hoarding Disorder," and implement tactical practices like bounded domains, episodic retrieval, and the Research→Plan→Reset→Implement cycle.

**Key Distinction:** Context stuffing assumes volume = quality ("paste the entire PRD"). Context engineering treats AI attention as a scarce resource and allocates it deliberately.

This is not about prompt writing—it's about **designing the information architecture** that grounds AI in reality without overwhelming it with noise.

## Input

**Works best with:** A description of the AI workflow, agent, or prompt setup that feels bloated, brittle, or hard to steer.
**Also useful:** What you've already stuffed into context (docs, transcripts, schemas) and where outputs go wrong.

Anything supplied with the invocation itself — text after the skill name, a pasted context dump, or an appended `ARGUMENTS:` line — counts as answers already given. Use it and skip whatever it covers; don't re-ask.

**Arriving empty-handed? That works too.** The advisor opens by asking what you're feeding the model today and what breaks.

**Example invocation:** `Diagnose my setup: our support-triage agent gets the full 40-page policy manual per ticket and still misroutes edge cases.`

## Key Concepts

### The Paradigm Shift: Parametric → Contextual Intelligence

**The Fundamental Problem:**
- LLMs have **parametric knowledge** (encoded during training) = static, outdated, non-attributable
- When asked about proprietary data, real-time info, or user preferences → forced to hallucinate or admit ignorance
- **Context engineering** bridges the gap between static training and dynamic reality

**PM's Role Shift:** From feature builder → **architect of informational ecosystems** that ground AI in reality

---

### Context Stuffing vs. Context Engineering

| Dimension | Context Stuffing | Context Engineering |
|-----------|------------------|---------------------|
| **Mindset** | Volume = quality | Structure = quality |
| **Approach** | "Add everything just in case" | "What decision am I making?" |
| **Persistence** | Persist all context | Retrieve with intent |
| **Agent Chains** | Share everything between agents | Bounded context per agent |
| **Failure Response** | Retry until it works | Fix the structure |
| **Economic Model** | Context as storage | Context as attention (scarce resource) |

**Critical Metaphor:** Context stuffing is like bringing your entire file cabinet to a meeting. Context engineering is bringing only the 3 documents relevant to today's decision.

---

### The Anti-Pattern: Context Stuffing

**Five Markers of Context Stuffing:**
1. **Reflexively expanding context windows** — "Just add more tokens!"
2. **Persisting everything "just in case"** — No clear retention criteria
3. **Chaining agents without boundaries** — Agent A passes everything to Agent B to Agent C
4. **Adding evaluations to mask inconsistency** — "We'll just retry until it's right"
5. **Normalized retries** — "It works if you run it 3 times" becomes acceptable

**Why It Fails:**
- **Reasoning Noise:** Thousands of irrelevant files compete for attention, degrading multi-hop logic
- **Context Rot:** Dead ends, past errors, irrelevant data accumulate → goal drift
- **Lost in the Middle:** Models prioritize beginning (primacy) and end (recency), ignore middle
- **Economic Waste:** Every query becomes expensive without accuracy gains
- **Quantitative Degradation:** Accuracy drops below 20% when context exceeds ~32k tokens

**The Hidden Costs:**
- Escalating token consumption
- Diluted attention across irrelevant material
- Reduced output confidence
- Cascading retries that waste time and money

---

### Real Context Engineering: Core Principles

**Five Foundational Principles:**
1. **Context without shape becomes noise**
2. **Structure > Volume**
3. **Retrieve with intent, not completeness**
4. **Small working contexts** (like short-term memory)
5. **Context Compaction:** Maximize density of relevant information per token

**Quantitative Framework:**
```
Efficiency = (Accuracy × Coherence) / (Tokens × Latency)
```

**Key Finding:** Using RAG with 25% of available tokens preserves 95% accuracy while significantly reducing latency and cost.

---

### The 5 Diagnostic Questions (Detect Context Hoarding Disorder)

Ask these to identify context stuffing:

1. **What specific decision does this support?** — If you can't answer, you don't need it
2. **Can retrieval replace persistence?** — Just-in-time beats always-available
3. **Who owns the context boundary?** — If no one, it'll grow forever
4. **What fails if we exclude this?** — If nothing breaks, delete it
5. **Are we fixing stru

SAFEgrade B · trust 89/100 Nothing in the source contradicts what it says it does. Grade A is reserved for packages that have also passed the behavioural sandbox.

What the source does

Filesystem

none-observed

Network

none-observed

Shell

none-observed

Dependencies

pinned

Secrets in source

none-found

Findings (0)

No findings outside the package's declared scope.

Gates applied: no_behavioural_pass.