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.
git clone --depth 1 https://github.com/sungmanch/claude-glm-agentWrote 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/commands/sungmanch/claude-glm-agent/task-with-glm)<a href="https://agentmods.dev/commands/sungmanch/claude-glm-agent/task-with-glm"><img src="https://agentmods.dev/badge/commands/sungmanch/claude-glm-agent/task-with-glm.svg" alt="Measured on agentmods" height="20"></a>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 | $0.00016 | $0.06082 |
| Opus 5 | $0.00008 | $0.03041 |
| Sonnet 5 | $0.00003 | $0.01216 |
| Haiku 4.5 | $0.00002 | $0.00608 |
Grade A, and why
task-with-glm 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.
How it starts
The opening of the file, as written. The whole thing — 887 lines — stays where its author put it; the contents beside it link to each section on GitHub.
[GLM TASK MODE ACTIVATED]
Task: $ARGUMENTS
Phase 0: Scenario Detection
Analyze the task using the keyword matrix below. If multiple scenarios match, select the PRIMARY scenario based on the user's core intent.
Scenario Detection Matrix
| Scenario | Detection Keywords |
|---|---|
| Bug/Debugging | error, bug, fix, crash, exception, not working, fails, broken, issue, debug, trace, stack, traceback, undefined, null |
| Code Review | review, check, evaluate, quality, feedback, critique, assess, approve, PR, pull request, merge request |
| New Feature | add, implement, create, build, develop, new, feature, functionality, capability, introduce |
| Refactoring | refactor, restructure, improve, clean up, simplify, optimize structure, redesign, rewrite, modernize |
| Exploration | understand, explain, how does, what is, analyze, architecture, overview, trace, flow, structure, codebase |
| Performance | slow, performance, optimize, speed, memory, latency, bottleneck, profile, efficient, cache |
| Testing | test, coverage, edge case, unit test, spec, mock, assertion, TDD, validation, e2e |
| Documentation | document, docs, README, comment, API docs, explain, JSDoc, docstring, describe, annotate |
| Research | research, look up, find information, documentation, library, API, how to use, learn about, context7, web search, latest, current, find docs, framework, package, best practices |
| Planning/Debate | plan, decide, choose, compare, trade-off, pros cons, should we, which approach, strategy, architecture decision, design decision, evaluate options, versus, vs, alternative |
Default: If no clear match, use General scenario (original behavior).
Phase 1: Skip Criteria Check
Skip GLM workers and use Opus alone when:
- Simple file read/write operations
- Git commands (commit, push, status)
- Single file edits under 50 lines
- User requests "quick" or "fast" response
- Insufficient code context provided
- Trivial formatting or linting fixes
- Direct questions with obvious answers
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 · 887 lines · 16 tokens per session scan A 3c726cca9f85
task-with-glm is a command published in the GitHub repository sungmanch/claude-glm-agent (5 stars, last pushed 5mo ago), licensed MIT. It adds 16 tokens to every session and 6,082 once invoked, about $0.0001 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 commands, from other repositories
awesome-chatgpt
Search awesome-ChatGPT-repositories for open-source GitHub repositories related to ChatGPT and LLMs.
commit
A command that examines staged Git changes and proposes a commit message using the Conventional Commits format, such as feat, fix, or docs. Git is a tool for tracking changes to code.
init
Scaffold a new MindBase project (v2 layout). Usage: /mb:init [template] [-- mission ...].
pr
Handle the full workflow from current branch state to an open, CI-monitored pull request.
recall
Search past tool-call observations with BM25 ranking to answer "have I seen this before?" before re-deriving a fix or repeating a mistake. Enforces Law 1 (Research Before Executing).
requirement-review
A command that has several specialized agents review a requirements document and produce a revised version through a controlled workflow.