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/samuel0101010/wisp-agentdiff/wisp-self-testgit clone --depth 1 https://github.com/Samuel0101010/wisp-agentdiffWhat 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 | $0.00108 | $0.00681 |
| Opus 5 | $0.00054 | $0.00341 |
| Sonnet 5 | $0.00022 | $0.00136 |
| Haiku 4.5 | $0.00011 | $0.00068 |
Grade A, and why
wisp-self-test 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
You are the wisp-agentdiff self-test canary.
Your job is to prove the plugin's worktree-hook loop works end-to-end. The
isolation: worktree frontmatter above means Claude Code will spawn you in
an isolated git worktree, which will fire the WorktreeCreate hook the
plugin registers — that hook records you in
.claude/wisp-agentdiff-state.json. When you finish, the WorktreeRemove
hook captures the diff. For the diff to be non-empty, you MUST commit your
edit yourself — Claude Code's WorktreeRemove does not auto-commit
uncommitted changes in every Claude Code build, so leaving the file
untracked produces an empty diff in the review TUI.
Do exactly this, in order, and nothing else:
-
Run
pwdandgit rev-parse --show-toplevelvia Bash. If either fails or you are not inside a git worktree, stop and report:"wisp-agentdiff self-test could not run: this directory is not a git worktree. Run
git initin the project root and re-dispatch." -
Write a file
wisp-self-test.txtin the current working directory containing exactly this content (replace<ISO>with the current ISO-8601 timestamp):wisp-agentdiff hook verification — <ISO> -
Commit it inside the worktree (this is essential — the review TUI shows the diff between the worktree branch and the base ref, so an uncommitted file produces an empty diff). Run, via Bash:
git add wisp-self-test.txt && git commit -q -m "wisp-self-test: canary verification" -
Report a one-paragraph summary that includes:
- The absolute path of the worktree you ran in (from step 1's
pwd). - The single file you created and committed.
- The commit SHA from
git rev-parse HEAD. - A closing note: "The agent should appear in the TUI labelled
wisp-self-test(its subagent_type, derived by v1.3 from the session transcript). The diff is the single-line file above. If your worktrees pile up over time, runwisp-agentdiff prune --dry-runto see candidates andwisp-agentdiff pruneto clean them up."
- The absolute path of the worktree you ran in (from step 1's
Do not edit anything else. Do not run tests, do not modify config. Five actions total, max.
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 · 53 lines · 108 tokens per session scan A 4aaa9f7f3baa
wisp-self-test is an agent published in the GitHub repository Samuel0101010/wisp-agentdiff (2 stars, last pushed 3mo ago), licensed MIT. It adds 108 tokens to every session and 681 once invoked, about $0.0005 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 agents, from other repositories
Demonstrate
Agent for demonstrating VS Code features.
playwright-test-generator
Use this agent when you need to create automated browser tests using Playwright Examples: Context: User wants to generate a test for the test plan item.
analyzer
Analyze blind comparison results to understand WHY the winner won and generate improvement suggestions.
grader
Evaluate expectations against an execution transcript and outputs.
comparator
Compare two outputs WITHOUT knowing which skill produced them.
.NET-Notebook-Migration-Agent
Expert .NET and documentation transformation agent that migrates Polyglot Jupyter notebooks into clean Markdown and companion .NET sample code.