- Home
- Skills
- Agents & Orchestration
- Buddie Discovery
Works with the AI Agents you already use
Buddie Discovery
Evidence triage for AI agents: use real Buddie Core verification when connected, check supplied or accessible evidence independently otherwise, and clearly mark unavailable checks.
Free
Buddie Discovery
Example session with this skill installed
Verify that the source says “All 41 tests passed.” Source text supplied here: “Results: 40 tests passed; 1 failed.” Also confirm PR999 merged with green CI; no PR evidence is provided.
- Read your context and instructions
- Compiled the buddie discovery
The supplied text contradicts “All 41 tests passed.” It says “Results: 40 tests passed; 1 failed.”
- Method: Direct comparison of the text you supplied.
- Finding: Contradicted by the supplied evidence.
- Limit: This compares wording only; it does not establish that any tests actually ran or authenticate the source.
PR #999 merged: not checked. No repository, PR record, or merge evidence was provided.
PR #999 has green CI: not checked. No check results tied to the relevant commit were provided.
To check the PR claims, provide its repository and PR evidence, plus CI results for the relevant commit. No external tools, network access, or Buddie Core were used; no Core receipt was generated.
Connects securely to your tools. The creator never sees your data.
About this skill
Buddie Discovery is the portable evidence-triage skill from Buddie — AI Agent Reliability Toolkit.
Verify the claim. Know the context.
It helps an agent separate checkable claims, identify the evidence each needs, and report findings with sources, revisions and limits. Use it for quotations, repository facts, PR/CI status, command-output claims and permitted environment diagnostics.
Three operating modes:
- Core-connected: discover the actual available Buddie MCP or CLI verifier, invoke it, and report its real result and scope.
- Standalone: compare evidence using tools already authorized in the session; label findings independently checked, contradicted, or not checked.
- No relevant tools: identify missing evidence and leave claims unverified.
The ZIP contains one canonical SKILL.md, references/reporting.md and the Apache-2.0 license. It contains no executable verifier, installer or bundled dependencies.
Buddie Core is optional. Formal Core receipts require a real Core invocation; independent observations are never presented as Core verdicts. Available tools, hooks, transcript checks and diagnostics vary by client. No automatic installation, privilege grants or private-evidence sharing. Projects/Context is not included.
The demo below is an actual instruction-following smoke using supplied text and no external tools. It is not a Claude platform certification.
Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for Agents & Orchestration.
Discovery skill version 0.1.0. No executable Core verifier is included. External claims stay unverified without access to evidence. Independent checks are not Core receipts. Client-specific hooks and diagnostics vary.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before buying a skill.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Free to use
No account required to browse
Trust & safety
Security scan
Security scan passed 19 days ago
- Free with an Agensi account
Nothing to install. It runs in your AI agent.
- Optional
- Buddie Core
Buddie Core is only for Core receipts.
Permissions
No mandatory account connections. Uses only tools and evidence already authorized for the task. Optional browser/network access retrieves sources; optional local file/terminal access checks local evidence. Environment diagnostics require Core and its existing grants. No permission is granted by this skill.