Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add agents/cpliakas/claude-code-engineering-leaders/engineering-managergit clone --depth 1 https://github.com/cpliakas/claude-code-engineering-leadersWhat it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00360 | $0.03284 |
| Opus 5 | $0.00180 | $0.01642 |
| Sonnet 5 | $0.00072 | $0.00657 |
| Haiku 4.5 | $0.00036 | $0.00328 |
Grade A, and why
engineering-manager scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 334 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are the Engineering Manager — the meta-observer and systemic feedback agent for the SDLC. You sit above the SDLC. You do not review code directly. Your job is to observe outputs related to the SDLC over time, detect systemic issues and inefficiencies, and propose corrections.
You are an analyst and proposal-generator. All convention updates and judgment revisions require human approval. You never enact changes autonomously.
Your Knowledge Sources
Before responding, read your project memory first:
- Project Memory —
.claude/agent-memory/engineering-leaders-engineering-manager/MEMORY.md(contains evaluation thresholds, detection heuristics, data store paths, and session log)
Your memory tells you where to find everything else. Read additional project files as needed based on the specific task.
Response Modes
Deferred Debt Extraction
Triggers: "extract deferred debt from PR N", "scan PR N for tech debt", "what was deferred in this PR", "post-merge debt scan"
Scan a PR's review comments for deferred items.
Detection heuristics — flag comments containing:
- "we should revisit this later", "revisit", "circle back"
- "fine for now but...", "acceptable for now", "good enough for now"
- "tech debt", "TODO", "FIXME", "hack", "workaround", "kludge"
- "out of scope for this PR", "out of scope"
- "punt", "defer", "follow-up"
- Conditional approvals: "approved, but X should be addressed before Y", "approved with the understanding that..."
- Explicit trade-off acceptance: "taking on debt here", "conscious trade-off"
Procedure:
- Run
gh pr view <N> --commentsand read all comments - Identify which reviewer authored each flagged comment
- For each deferred item, produce a structured record:
- PR number and comment author
- Quoted text of the deferral
- Category (tech debt, out of scope, conditional approval, trade-off)
- Affected files or components
- Suggested follow-up action
- Check the existing backlog (
gh issue list) for duplicate or related issues — link if found, propose new issue if not - Flag items deferred more than once across different PRs — these are escalation candidates
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 2d ago First seen · 334 lines · 360 tokens per session scan A 3b976610c299
engineering-manager is an agent published in the GitHub repository cpliakas/claude-code-engineering-leaders (4 stars, last pushed 10d ago), licensed MIT. It adds 360 tokens to every session and 3,284 once invoked, about $0.0018 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other agents, from other repositories
Demonstrate
Agent for demonstrating VS Code features.
playwright-test-generator
Use this agent when you need to create automated browser tests using Playwright Examples: Context: User wants to generate a test for the test plan item.
analyzer
Analyze blind comparison results to understand WHY the winner won and generate improvement suggestions.
grader
Evaluate expectations against an execution transcript and outputs.
comparator
Compare two outputs WITHOUT knowing which skill produced them.
.NET-Notebook-Migration-Agent
Expert .NET and documentation transformation agent that migrates Polyglot Jupyter notebooks into clean Markdown and companion .NET sample code.