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 skills add 0xranx/golembot --skill escalationgit clone --depth 1 https://github.com/0xranx/golembotWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/0xranx/golembot/escalation)<a href="https://agentmods.dev/skills/0xranx/golembot/escalation"><img src="https://agentmods.dev/badge/skills/0xranx/golembot/escalation.svg" alt="Measured on agentmods" height="20"></a>- NVIDIA SkillSpector pass
What 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 5.1 | $0.00066 | $0.00446 |
| Opus 5 | $0.00033 | $0.00223 |
| Sonnet 5 | $0.00013 | $0.00089 |
| Haiku 4.5 | $0.00007 | $0.00045 |
Grade A, and why
escalation 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 8d 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.
What it actually says
Human Escalation Protocol
When you encounter a situation you cannot handle, escalate to a human agent by recording the escalation.
When to Escalate
- You cannot confidently answer the user's question
- The user explicitly asks to speak to a human
- The request involves sensitive operations (financial, legal, security)
- You detect a safety concern
- The user is frustrated after multiple failed attempts
How to Escalate
- Inform the user that their request is being escalated
- Write an escalation record to
.golem/escalations.jsonl(one JSON object per line):
echo '{"ts":"2026-03-15T10:00:00Z","sessionKey":"feishu:chat123:user456","reason":"User requested human support for billing issue","context":"User asked about refund policy, I could not find the answer","status":"open"}' >> .golem/escalations.jsonl
Escalation Record Schema
| Field | Type | Description |
|---|---|---|
ts |
string | ISO 8601 timestamp |
sessionKey |
string | The current session key |
reason |
string | Why this is being escalated |
context |
string | Brief summary of what was discussed |
status |
string | Always "open" when creating |
Response Template
When escalating, respond to the user like:
I've flagged this for human review. A team member will follow up on your request about [topic]. In the meantime, is there anything else I can help with?
Verifying Escalation
After writing the record, confirm the file exists and the entry was appended:
tail -1 .golem/escalations.jsonl
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.
- 8d ago First seen · 50 lines · 66 tokens per session scan A ef7261cff0af
escalation is a skill published in the GitHub repository 0xranx/golembot (320 stars, last pushed 2d ago), licensed MIT. It adds 66 tokens to every session and 446 once invoked, about $0.0003 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 2026-08-30.
Other skills, from other repositories
karpathy-guidelines
Tool-agnostic behavioral guidelines for AI coding assistants. Use when writing, reviewing, debugging, or refactoring code to reduce overengineering, surface ambiguity, make surgical changes, and define verifiable success criteria.
ultracite
Ultracite is a zero-config linting and formatting preset for JavaScript/TypeScript projects. Use when: (1) Setting up or initializing Ultracite in a project (ultracite init), (2) Running linting or formatting commands (check, fix, doctor), (3) Writing or reviewing JS/TS code in a project that uses Ultracite — to…
ulw-execute
Executes a written Prometheus work plan with Boulder state, evidence ledger, worktree discipline, and parallel subagents. Use when the user says ulw-execute or asks to run a .omo/plans plan.
review-work
Post-implementation gate review: run manual QA on the real surface yourself, then launch ONE gate reviewer (never a panel) to audit goal, constraints, code quality, security, missed context, and QA evidence. Use before a PR handoff or when the user explicitly asks to review completed work.
remove-deadcode
Remove unused code from this project with ultrawork mode, LSP-verified safety, atomic commits. Triggers: remove dead code, dead code, cleanup, remove unused.
ulw-loop
A goal-like loop that decomposes work into systematic, evidence-bound ultrawork steps. Use when the user wants a goal loop or durable, checkpointed execution.