A production API
A 43-operation API graded A, still missing what no spec can carry
Pegana is real stablecoin infrastructure on Solana: peg state, depeg simulation, loop exposure, webhook keys, 43 operations across 39 paths, in production, used by real money. Run through our agent-readiness scorecard it grades A, 100 out of 100. This isn't a list of what Pegana got wrong. It's what no specification can carry, at any version, at any quality, which is exactly why a good API is the honest place to show it
operations
scorecard grade
context compression
“A specification tells you what a call looks like. It cannot tell you whether it will work, and that's structural, not a defect anyone can fix by writing better docs”
Not the model. Not the agent. Not Pegana's API, which grades A. What the call was built from
"qual o peg state do USDC?" resolved correctly and came back: USDC pegged, $0.99975
"Is the stablecoin in this headline still pegged, and what is it trading at?" spans three surfaces. The join across all three is confirmed against real responses, offline, keyless, $0
Measured 2026-08-06, not estimated. Shrinking a spec is easy if you're willing to drop things; holding correctness while you do it is the part that's worth anything
WHEN THIS APPLIES TO YOU
This is the case for a provider deciding whether comprehension is worth doing on a spec that already grades well. A high scorecard doesn't mean nothing is left to fix, it means what's left isn't fixable by writing better docs. That's the argument for doing this work even on your best-documented endpoint
mcp.geckovision.tech/pegana/mcp
