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 skills/adigoel07/tooloftruth/skillnpx skills add adigoel07/tooloftruth --skill skillgit clone --depth 1 https://github.com/adigoel07/tooloftruthWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/adigoel07/tooloftruth/skill)<a href="https://agentmods.dev/skills/adigoel07/tooloftruth/skill"><img src="https://agentmods.dev/badge/skills/adigoel07/tooloftruth/skill.svg" alt="Measured on agentmods" height="20"></a>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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00073 | $0.00683 |
| Opus 5 | $0.00036 | $0.00342 |
| Sonnet 5 | $0.00015 | $0.00137 |
| Haiku 4.5 | $0.00007 | $0.00068 |
Grade A, and why
tooloftruth 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.
How it starts
The opening of the file, as written. The whole thing — 83 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Tool of Truth
Every tool call, proven.
What This Does
Tool of Truth verifies that AI agents actually used the tools they claim to use. It prevents tool-use hallucination — where agents fabricate tool usage instead of actually calling the tool.
Verification Protocol (MANDATORY)
Before claiming ANY tool was used:
Step 1: Check if the tool is available
- Run:
which <tool>ORnpm ls <tool>ORpip show <tool> - If not found → STOP. Tell user the tool is not installed.
Step 2: Check if you actually called it
- Review your own execution: did you make an API call, CLI invocation, or MCP tool call?
- If you only READ about the tool (docs, SKILL.md) but didn't CALL it → you did NOT use it. Say so.
Step 3: Verify the output is real
- Does the output contain real data from the tool?
- Or does it look like it was generated from memory/training?
- If output is suspiciously perfect or matches the docs exactly → flag it as potentially fabricated.
Rules
- NEVER claim a tool was used without evidence
- If the tool is not available, SAY SO immediately
- If you're unsure whether you used it, verify before claiming
- Include verification status in responses when Tool of Truth MCP is connected
- If fabrication is detected, STOP and report to user
- LOG EVERY CLAIM: Before claiming you used a tool, call
tooloftruth_log_claimwith the tool name and what you're about to say. This creates an audit trail for honesty verification. - AUDIT YOURSELF: Periodically call
tooloftruth_auditto check if your claims match reality.
Using /truth
Type /truth to verify the last tool interaction.
Type /truth <tool-name> to verify a specific tool.
The command returns:
- Whether the tool was actually used
- Trust score (0–100)
- A verification receipt
Connecting Tool of Truth MCP (Recommended)
For maximum verification accuracy, connect the Tool of Truth MCP server:
npm install -g tooloftruth-mcp
Add to your MCP config:
"tooloftruth": { "command": "tooloftruth-mcp" }
What ships with it
1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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 · 83 lines · 73 tokens per session scan A 02e4bc09855d
tooloftruth is a skill published in the GitHub repository adigoel07/tooloftruth (0 stars, last pushed 16d ago), licensed MIT. It adds 73 tokens to every session and 683 once invoked, about $0.0004 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 skills, from other repositories
aep
Use this skill whenever working with AEP (Agent Element Protocol) 2.8, dynAEP (main AEP event runtime), Base Node, Composer Lite, CCA / setup agent, component registry, CAW, UCB, Path A/B connect, dynAEP-TA, dynAEP-TA-P or any AEP governance feature. Triggers include 'AEP', 'dynAEP', 'dynAEP-TA', 'dynAEP-TA-P'…
aep-caw-policy-create
Use when creating a new AepCaw security policy, including agent sandboxes, CI pipelines, development environments, HTTP service gateways, or Postgres-family database access policies.
aep-caw-policy-edit
Use when adding, removing, or updating rules in an existing AepCaw policy, modifying security permissions, HTTP service declarations, Postgres-family database rules, resource limits, or policy YAML files.
eval-env
dangerous eval pattern.
git-origin
Has a source frontmatter field.
minimal
A minimal valid skill.