File	What it is
`01_valid_receipt.json`	A valid signed receipt for a `lookup_flights` tool call. Verification returns True.
`02_tampered_receipt.json`	The same receipt with one field modified after signing. Verification returns False.
`03_chain_three_receipts.json`	A chain of three valid receipts (search, hold, book) with `previous_receipt_hash` linking each to the prior one.

Verifying the samples

The notebook walks through verification in four sections. To verify these fixtures directly without running through the notebook narrative:

import json
from pathlib import Path

# Assumes you have completed the imports and helper functions
# from sections 1 and 2 of 18-signed-receipts.ipynb.

valid = json.loads(Path("01_valid_receipt.json").read_text())
print(f"Valid receipt: {verify_receipt(valid)}")        # True

tampered = json.loads(Path("02_tampered_receipt.json").read_text())
print(f"Tampered receipt: {verify_receipt(tampered)}")  # False

chain = json.loads(Path("03_chain_three_receipts.json").read_text())
for r in verify_chain(chain):
    print(f"  Receipt {r['index']} ({r['tool']}): {'VALID' if r['overall_valid'] else 'INVALID'}")

How these were generated

The fixtures use the same code path as the notebook, with one fixed signing key and fixed timestamps for byte-reproducibility. To regenerate:

python3 generate_fixtures.py

(Script is at generate_fixtures.py in this directory.)

What students learn from inspecting raw JSON

Reading the raw receipt format builds intuition that the cells in the notebook do not always provide. Students who skim the JSON often notice:

The signature is an opaque base64url string, but every other field is plain readable JSON. The signature does not encrypt the content; it attests to it.
The public_key is embedded in the receipt. An auditor needs nothing else to verify (subject to trusting that the key actually belongs to the claimed issuer; see the lesson README on identity infrastructure).
Modifying a single character of any field, then re-comparing this file with 02_tampered_receipt.json, makes the byte-level mechanism concrete.