Minimal, runnable agents that record verifiable proof of work on MINT after completing a task — one per framework. Each one:
- registers a MINT identity on first run and caches it (
~/.mint/<agent>.json), - does one simple task,
- attests the result, and
- prints the attestation and your running work count.
No wallet, no keys, no blockchain code. Set MINT_API_KEY=fnet_… to attest under
your own account; otherwise each agent self-provisions a free scoped identity and
runs on the free daily cap — so they work out of the box with no signup.
| Example | Install | Runs with no LLM key? |
|---|---|---|
langchain_attesting_agent.py |
pip install mint-attest langchain |
✅ yes |
autogen_attesting_agent.py |
pip install mint-attest pyautogen |
✅ yes (set OPENAI_API_KEY for a real chat turn) |
crewai_attesting_agent.py |
pip install mint-attest[crewai] crewai |
✅ yes (set OPENAI_API_KEY for a real crew) |
llamaindex_attesting_agent.py |
pip install llama-index-tools-mint llama-index llama-index-llms-openai |
✅ yes (set OPENAI_API_KEY for a real agent) |
# e.g.
pip install mint-attest langchain
python langchain_attesting_agent.pyEach prints a MINT identity: MINT-xxxxxx and an attestation; check your agent's
work history at mint-explorer.vercel.app.
See ../AGENT_PROMPT_SNIPPET.md for copy-paste prompt
and code blocks, and ../INTEGRATION.md for the full guide
(payment flow, FAQ). Semantic Kernel users: see
mint-semantic-kernel.