Trust Log Skill#

Use trustlog to produce local JSON + Markdown receipts for agent work. Receipts should make it easy for a human or next agent to see what ran, what changed, what looked risky, and what was redacted.

Default workflow#

1. From the target project root, wrap meaningful verification commands:

bash

npx @builtbyecho/trustlog run -- npm test
npx @builtbyecho/trustlog run -- npm run build

1. Read the human summary:

bash

cat .trustlog/latest.md

1. Verify before sharing or using as final evidence:

bash

npx @builtbyecho/trustlog verify .trustlog/latest.json

When to use#

• Final proof after tests/builds/lints.
• Handoff evidence for another agent.
• User asks “what happened?” or “give me a receipt.”
• Risky commands where a durable local audit trail helps.

Safety#

• Prefer non-destructive commands.
• Do not use this to justify commands you would not otherwise run.
• Receipts redact common secrets and strip thinking-looking blocks, but still inspect before posting publicly.
• If verification fails, do not share the receipt until the issue is fixed or clearly explained.

Useful commands#

bash

npx @builtbyecho/trustlog run -- <command> [args...]
npx @builtbyecho/trustlog summarize .trustlog/latest.json
npx @builtbyecho/trustlog verify .trustlog/latest.json

Vaultline integration#

After verification passes, upload the trust receipt so reviewers can access one durable artifact instead of local file paths.

bash

node -e '
const fs = require("node:fs");
const { Vaultline } = require("@builtbyecho/vaultline-sdk");
const receiptPath = ".trustlog/latest.md";
const content = fs.readFileSync(receiptPath, "utf8");
const vault = new Vaultline({ apiKey: process.env.VAULTLINE_API_KEY });
vault.files.upload({
  path: `trustlog/${Date.now()}-latest.md`,
  content,
  contentType: "text/markdown"
}).then(r => console.log("Vaultline fileId:", r.fileId));
'

Use this when proof needs to be preserved across agents, channels, or time.