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/taoidle/plan-cascade/autogit clone --depth 1 https://github.com/Taoidle/plan-cascadeWhat 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 | $0.00041 | $0.08637 |
| Opus 5 | $0.00020 | $0.04319 |
| Sonnet 5 | $0.00008 | $0.01727 |
| Haiku 4.5 | $0.00004 | $0.00864 |
Grade A, and why
auto 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 yesterday.
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 — 939 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Plan Cascade - Auto Strategy Executor
AI automatically analyzes the task and executes the optimal strategy without user confirmation.
CRITICAL CONSTRAINTS (READ FIRST)
NEVER use EnterPlanMode. This command has its own planning and execution flow. Using Claude's native plan mode bypasses Plan Cascade entirely and defeats the purpose of this command. If you feel the urge to "enter plan mode" or "design an implementation approach", STOP — that is exactly what the Skill tool routing in Step 5 handles.
NEVER use the Task tool with subagent_type=Explore or subagent_type=Plan during this command's execution. Context gathering in Step 2 must use only Glob, Read, Grep, and Bash (for git commands only). Launching exploration agents delays routing and causes fallthrough to native plan mode.
After completing Step 3 (strategy analysis), you MUST proceed directly to Step 5 (Skill tool routing). Do not pause to "think about the approach", "gather more context", or "design a solution". The specialized skill invoked in Step 5 handles all of that.
Command-Line Flags
The auto command supports the following flags to customize execution:
Parameter Priority
When parameters are specified in multiple places, the following priority order applies:
- Command-line flags (highest priority) - Override all other sources
- Configuration files (prd.json, mega-plan.json) - Used when no command-line flag provided
- Default values (lowest priority) - Used when no other source provides the parameter
Example:
# If prd.json contains: flow_config.level = "standard"
# And you run: /plan-cascade:auto --flow full "task"
# Result: FLOW = "full" (command-line overrides configuration)
Note: Parameters are propagated through the execution chain:
auto→ passes parameters tohybrid-auto/hybrid-worktree/mega-plan- These commands → save parameters to
prd.json/mega-plan.json approve→ reads parameters from configuration files (can be overridden by command-line)
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.
- yesterday First seen · 939 lines · 41 tokens per session scan A 05f9a46c9d0d
auto is a command published in the GitHub repository Taoidle/plan-cascade (124 stars, last pushed 5mo ago), licensed MIT. It adds 41 tokens to every session and 8,637 once invoked, about $0.0002 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
conv_release
Command "conv_release" from John-Wendell/Attention-MoA, covering chatbot arena conversations, all conversations and prompt distribution.
data_cleaning
Command "data_cleaning" from John-Wendell/Attention-MoA, covering data cleaning, requirements, steps, convert html to markdown and keep or remove specific languages.
pypi
Command "pypi" from John-Wendell/Attention-MoA, covering requirement and upload.
local_cluster
node-01.
webserver
Command "webserver" from John-Wendell/Attention-MoA, covering install, launch servers, check the launch time, increase the limit of max open files and gradio edit (3.35.2).
leaderboard
Command "leaderboard" from John-Wendell/Attention-MoA, covering get logs, clean battle data, run elo analysis, copy files to hf space and update files on webserver.