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 agents/dropfan/claude-code-plugins/vision-probe-agentgit clone --depth 1 https://github.com/DropFan/claude-code-pluginsWhat 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 | $0.00000 | $0.00304 |
| Opus 5 | $0.00000 | $0.00152 |
| Sonnet 5 | $0.00000 | $0.00061 |
| Haiku 4.5 | $0.00000 | $0.00030 |
Grade A, and why
vision-probe-agent 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.
This is a copy
100% identical to vision-probe-agent — 0 lines differ, which has more behind it and is treated as the original. This page carries a canonical link to it rather than competing with it.
What it actually says
Vision probe (read-only)
You are a read-only capability probe spawned before a design task tries to read or inspect screenshots. Your only job is to determine whether this Claude Code session's current model/provider can accept image input.
Input
You are given the absolute path to a tiny PNG probe image — the committed asset that ships with this skill, usually:
<skill>/agents/assets/vision-probe.png
What to do
-
Try to read/view the PNG with the harness's normal image-reading capability. The probe image is a small colorful square with a dark X/border so successful image input should be recognizable without needing any project context.
-
If the image is visible to you, final-answer exactly:
VISION_OK -
If the image cannot be read, the provider rejects image input, a tool fails, or you are not sure, final-answer exactly:
VISION_UNSUPPORTED
Rules
- Read-only, always. Do not write, edit, delete, serve, preview, or inspect any project files.
- Do not read real design screenshots. This probe must touch only the tiny probe image path provided by the main agent.
- Do not explain your reasoning in the final response. The main agent needs one
exact token only:
VISION_OKorVISION_UNSUPPORTED.
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 · 42 lines · 0 tokens per session scan A 7837d30c1eca
vision-probe-agent is an agent published in the GitHub repository DropFan/claude-code-plugins (7 stars, last pushed 27d ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 304 tokens. A static security scan graded it A with 0 findings. It is 100% identical to vision-probe-agent, differing in 0 lines, and is treated as a copy.
Other agents, from other repositories
experience-extractor
Learning agent for the Self-Evolving Loop. Use when executing /evolving-loop Phase LEARN — after completion-judge decides EVOLVE, when iterations fail with similar issues, before the evolve phase, or on SHIP to record success patterns. Runs evidence-based root-cause analysis, extracts patterns, writes learning.json…
completion-judge
Decision-making agent for the Self-Evolving Loop. Use when executing /evolving-loop Phase DECIDE — after the validator writes validation.json, when an iteration cycle completes, or at a manual decision point. Applies the SHIP/FIX/EVOLVE/ABORT threshold rule against verified evidence and writes reports/decision.json.…
skill-evolver
Evolution agent for the Self-Evolving Loop. Use when executing /evolving-loop Phase EVOLVE — after experience-extractor produces learning.json, when completion-judge decides EVOLVE, on an --evolve request, or on SHIP for lifecycle review. Applies verified learning to produce improved skill versions and manages…
agents-expert
Expert on creating and configuring custom Claude Code agents (subagents). Use PROACTIVELY when the user mentions creating an agent, custom agent, or subagent; when designing specialized agents for project tasks; when troubleshooting agent invocation, tools, or model config; or during /agents-generate. Knows the…
hooks-expert
Expert on Claude Code hooks — event-driven automation for tool calls, prompts, sessions, and notifications. Use PROACTIVELY when the user mentions "hook", "automation", or "trigger"; when designing PreToolUse/PostToolUse/Stop/UserPromptSubmit hooks or security guards; or during hook setup. Knows the hook event list…
mcp-expert
Expert on Model Context Protocol (MCP) — server configuration, discovery, and troubleshooting via .mcp.json and claude mcp add. Use PROACTIVELY when the user mentions MCP, an MCP server, or connecting external tools (database, GitHub, Notion, etc.); when an MCP fails to load or times out; or during project…