avc

A visual planning workflow using AVC, a tool that displays structured plans in an interactive web view. People can review, reorder, edit, skip, or add steps before work begins.

In plain words
What is it for?
Use it for execution plans with more than three steps, architecture changes, multi-file refactors, and deployment sequences that need human review.
Why use it?
It makes complicated plans easier to inspect and approve than a long block of text. This is useful when a change spans several files, modules, or deployment steps.

Skill for Claude CodeCodex

Install

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.

agentmods
npx agentmods add skills/study8677/agent_view_controller-avc/avc
Any agent
npx skills add study8677/Agent_View_Controller-AVC --skill avc
Clone the repo
git clone --depth 1 https://github.com/study8677/Agent_View_Controller-AVC

Made for: Claude Code, Codex.

Per session 71 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,218 The whole file, excluding the scripts and references it only reads on demand.
Security scan B 1 finding. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5 $0.00071 $0.01218
Opus 5 $0.00036 $0.00609
Sonnet 5 $0.00014 $0.00244
Haiku 4.5 $0.00007 $0.00122

Measured 2d ago against content hash 0819cd7b4a18, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade B, and why

avc scanned grade B with 1 finding 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.

Asks for rootmediumPrivilege escalation

A mod that escalates privileges can change anything on the machine, not only the project.

cd Agent_View_Controller-AVC && go build -o avc . && sudo cp avc /usr/local/bin/
skills/avc/SKILL.md · 139 lines

How it starts

The opening of the file, as written. The whole thing — 139 lines — stays where its author put it; the contents beside it link to each section on GitHub.

AVC — Agent View Controller

AVC is a visual interaction tool that transforms JSON into interactive native WebView UIs. Use it when you need human approval for complex decisions.

When to Use

Use avc instead of printing plain text when:

  • Execution plans with more than 3 steps
  • Architecture changes involving multiple modules
  • Multi-file refactoring plans
  • Deployment sequences that need human review
  • Any complex plan where the human should visually review and reorder steps

Prerequisites

avc must be installed and available in PATH:

# From source
git clone https://github.com/study8677/Agent_View_Controller-AVC.git
cd Agent_View_Controller-AVC && go build -o avc . && sudo cp avc /usr/local/bin/

How to Use

Step 1: Construct JSON

Build a JSON object with view type and structured data:

{
  "view": "plan",
  "title": "Your Plan Title",
  "lang": "en",
  "editable": true,
  "token_count": 4500,
  "data": {
    "steps": [
      {"id": 1, "label": "Step description", "status": "pending"},
      {"id": 2, "label": "Another step", "status": "pending"}
    ]
  }
}

Tip — lang field: Localizes UI chrome (buttons / status bar / tooltips) only — not step content. Supported: en (default), zh, ja, ko, es, fr, de. Match it to the language of your step labels for a consistent UI.

Step 2: Pipe to AVC

echo '<json>' | avc

Step 3: Read the Result

  • Exit code 0 = human confirmed → stdout contains modified JSON
  • Exit code 130 = human cancelled
RESULT=$(echo '{"view":"plan","title":"My Plan","token_count":5000,"data":{"steps":[{"id":1,"label":"Step 1","status":"pending"}]}}' | avc)

if [ $? -eq 0 ]; then
  # Parse $RESULT — human may have reordered, edited, or removed steps
  echo "$RESULT"
else
  echo "Human cancelled"
fi

Important Behavior

  1. avc blocks until the human clicks Confirm or Cancel
  2. stdout contains the human-modified JSON (may have reordered/edited/removed steps)
  3. If token_count ≤ 3000 (default threshold), AVC passes through without opening a window
  4. Use --no-threshold to always show the WebView window
  5. Use --threshold=N to set a custom threshold

Read the full file on GitHub · 139 lines

Changes

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.

  1. 2d ago First seen · 139 lines · 71 tokens per session scan B 0819cd7b4a18

Subscribe to this mod's changes

avc is a skill published in the GitHub repository study8677/Agent_View_Controller-AVC (68 stars, last pushed 3mo ago), licensed Apache-2.0. It adds 71 tokens to every session and 1,218 once invoked, about $0.0004 per session on Opus 5. A static security scan graded it B with 1 finding (asks for root). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.