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 commands/logi-cmd/agent-guardrails/plangit clone --depth 1 https://github.com/logi-cmd/agent-guardrailsWhat 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 | $0.00011 | $0.00494 |
| Opus 5 | $0.00005 | $0.00247 |
| Sonnet 5 | $0.00002 | $0.00099 |
| Haiku 4.5 | $0.00001 | $0.00049 |
Grade A, and why
plan 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.
What it actually says
/ag:plan - Task Planning
Triggers
- User wants to plan a coding task
- User describes a feature or fix to implement
- User types
/ag:plan
Context Trigger Pattern
/ag:plan <task-description> [--intended-files <file1>, <file2>] [--risk-level standard|high]
Default behavior: npx agent-guardrails plan --task "<task-description>"
Behavioral Flow
- Confidence check: Assess confidence before planning
- ≥90%: Proceed directly
- 70-89%: Present plan alternatives
- <70%: Ask user for clarification
- Plan: Execute
npx agent-guardrails plan --task "<description>" - Review: Show task contract (allowed paths, intended files, risk level)
- Confirm: Ask user if plan looks correct before implementation
Key Patterns
- Scope definition: Clearly define what files are in/out of scope
- Risk assessment: Flag high-risk changes (auth, security, critical paths)
- Evidence tracking: Set up evidence file for task tracking
Examples
Basic planning
/ag:plan Add user authentication
# Runs: npx agent-guardrails plan --task "Add user authentication"
# Shows: task contract with allowed paths, risk level
Planning with scope
/ag:plan Fix login bug --intended-files src/auth.js, tests/auth.test.js
# Runs: npx agent-guardrails plan --task "Fix login bug" --intended-files src/auth.js tests/auth.test.js
High-risk planning
/ag:plan Refactor database layer --risk-level high
# Runs: npx agent-guardrails plan --task "Refactor database layer"
# Marks as high-risk, requires explicit confirmation
Boundaries
Will:
- Create task contract with clear scope
- Assess confidence before planning
- Show plan for user confirmation
Will Not:
- Start implementation without user approval
- Override user-specified constraints
- Proceed with low-confidence plans without alternatives
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 · 69 lines · 11 tokens per session scan A d043d15c8702
plan is a command published in the GitHub repository logi-cmd/agent-guardrails (29 stars, last pushed 17d ago), licensed MIT. It adds 11 tokens to every session and 494 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 2026-08-30.
Other commands, from other repositories
upgrade
Upgrade the skillshare CLI binary and/or the built-in skillshare skill.
build
Run full verification pipeline.
doctor
Run IJFW health check (files, MCP server, hooks, memory, caps, framing).
release
Walk the llmwiki release process step by step.
add-agent
引导新增一个 Agent 适配器。用法 /add-agent.
capture-feedback
Quick feedback capture with structured signals.