signafaide 0.0.0-stage → 0.2.1

package/README.md

@@ -1,3 +1,240 @@
-# Temporary Holding Version
+# signafaide
 
-This version is a temporary placeholder for this package. An operational version to replace this has been submitted for review and is awaiting a staged release.
+Signed, chained records of what your AI agents decided. Each record is signed with the agent's own key and chained on your machine; once you opt in, a witness (the signafAIde service, or one you host) co-signs and publicly timestamps it, so neither side can forge or edit a record alone.
+
+- Nothing is sent until a person runs `npx signafaide init` or sets `SIGNAFAIDE_API_KEY`. After that each record sends its content hash with its tool name, agent name and time (and its chain links), never its inputs or outputs unless you turn on stored content. Tool and agent names are sent as you write them, so don't put personal data in them. Each content hash is salted with 16 random bytes that stay on your machine with the content, so the hash alone cannot be used to guess the content, even when it has only a few possible values.
+- The sandbox `init` creates needs no account, email or card.
+- To keep the evidence in your environment, self-host the witness and set `SIGNAFAIDE_URL` ([self-host](https://signafaide.co.uk/docs/self-host)).
+- Apache-2.0, with a public protocol spec and an independent verifier.
+
+```bash
+npm install signafaide
+```
+
+```ts
+import { record } from "signafaide";
+
+const approvePayout = record(async (claimId: string, amount: number) => { /* ... */ }, { name: "approve_payout" });
+```
+
+```bash
+npx signafaide verify
+```
+
+That's it. With no API key, records are signed, chained and kept in `.signafaide/`, and nothing is sent. `signafaide verify` checks every signature, chain link and content hash and ends with a result line that says how many of the 12 checks ran, for example `PASS: 4 of 12 checks passed, 8 not checked (nothing co-signed yet)`. A check that was not run is never counted as passed. It also says what still needs the witness:
+
+```
+co-signature, public timestamp: not yet. Run `signafaide init` to send hashes, or self-host (SIGNAFAIDE_URL)
+```
+
+## Sending hashes to a witness
+
+Co-signature and the public timestamp need a witness. A person opts in once:
+
+- `npx signafaide init` creates a free sandbox (no account, email or card), saves `SIGNAFAIDE_API_KEY` to `.env`, adds `.env` to `.gitignore` and prints a link to claim it; or
+- set `SIGNAFAIDE_API_KEY` (a live key from the console), with `SIGNAFAIDE_URL` for a self-hosted witness. `SIGNAFAIDE_URL` alone sends nothing.
+
+Records written before the opt-in ship afterwards with their original times. The service accepts records up to 7 days old (an agent key's `valid_from` is at most 7 days before its registration); a local chain that starts earlier stays local, still verifies, and `signafaide flush` reports it as `too_old`. The SDK's `init()` only configures: it never creates a sandbox.
+
+## What a call records
+
+1. A `decision` record before the function runs: the name, plus `{"input": <the argument, or the argument list if there are several>, "meta": {...}}`.
+2. The function runs.
+3. An `outcome` record after it: `{"output": ...}`, or `{"error": {"type", "message"}}`. The error is thrown again unchanged.
+
+Records are signed and chained locally in `.signafaide/` (synchronously), then shipped in the background. Sync functions stay sync, and async functions stay async. Named functions don't need `name`.
+
+```ts
+const approve = record(approvePayout, { agent: "claims-bot", name: "approve_payout", meta: { model: "gpt-x" } });
+note("policy_loaded", { version: 3 }); // anything that is not a tool call
+```
+
+Give each agent an explicit name, for example `wrapTools(tools, { agent: "claims-bot" })`; without one it is the project folder's name. It must match `^[a-z0-9][a-z0-9._-]{0,63}$`. In production, one agent name is one chain written by one running instance.
+
+## Frameworks
+
+```ts
+import { wrapTools } from "signafaide";
+
+const tools = wrapTools({ approvePayout, escalate }); // object: key = name
+const list = wrapTools([lookupOrder, refund]);        // array
+```
+
+`wrapTools` (and `wrapTool` for one) also takes:
+
+- **Vercel AI SDK** tools (objects with `execute`). It records `execute(input)`; the options argument (messages, abort signal) is not recorded.
+  ```ts
+  const result = await generateText({ model, tools: wrapTools({ weather: tool({ inputSchema, execute }) }), prompt });
+  ```
+- **OpenAI Agents JS** function tools (`tool({...})` from `@openai/agents`). It records `invoke`.
+  ```ts
+  const agent = new Agent({ name: "support", tools: wrapTools([lookupOrder]) });
+  ```
+- **LangChain.js** tools (`tool()`, `DynamicStructuredTool`, `StructuredTool` subclasses). Every `invoke` is recorded, including tool calls, and the wrapped copy is still a `StructuredTool`.
+
+None of them is a dependency: `wrapTools` recognises the objects by their shape.
+
+## Strict mode: no receipt, no action
+
+```ts
+const wireFunds = record(sendWire, { strict: true }); // now returns a Promise
+```
+
+The decision is shipped first, and the function runs only after the service has returned its receipt. If the service can't be reached or refuses the record, or there is no API key (`no_key`), the call rejects with `NotRecorded` and the function does not run. Agent frameworks often turn a tool's error into a failed tool call and print nothing, so the first time a strict tool is refused in a process the SDK writes one warning line to stderr with the tool's name and the fix.
+
+## Hash-only
+
+By default the service receives only the envelope: the name, the time, the chain links and `content_hash`. The hash is `sha256(salt || JCS(content))` with a random 16-byte salt per record, kept next to the content in the local chain, so the service cannot confirm a guess of the content from the hash. When you store content with the service, the salt goes with it, so anyone holding the pack can check the content against the hash. Your inputs and outputs stay in `.signafaide/chain/<project>/<env>/<agent>.jsonl` (`.signafaide/chain/_pending/nokey/<agent>.jsonl` before the opt-in), which is your evidence store. To send the content too (encrypted at rest; live records need the Team plan):
+
+```ts
+await init({ storeContent: true });           // everything
+record(fn, { name: "x", storeContent: true }); // one tool
+```
+
+## Offline
+
+Nothing waits on the network unless you use strict mode. With an API key, records go to `.signafaide/outbox/` and are shipped in batches, one request at a time: every 2 seconds, every 100 records, and once more when the process is about to exit, for at most 3 seconds (`exitFlushMs`, `SIGNAFAIDE_EXIT_FLUSH_MS`). If the service can't be reached, they stay on disk and are retried with a backoff that grows to a minute; a refusal retrying cannot fix (quota used up, a wrong or revoked key) stops sending until the configuration changes. A malformed `SIGNAFAIDE_API_KEY` is caught here and never sent. They ship on the next run or with `signafaide flush`. Records are never lost on exit, and background requests never keep your process alive.
+
+## Containers that restart
+
+A container with no persistent disk continues its chain after the witness's head, and its first record is a signed start note. So that a routine restart is not read as a possible loss, end the instance with `await stop()`: it ships what is left and then writes a signed stop note. A start right after a stop note is a clean restart; a start with no stop note before it (a crash, a kill, an unreachable witness) is shown as a warning, because records the earlier instance had not sent are not known.
+
+```ts
+import { stop } from "signafaide";
+
+process.on("SIGTERM", async () => {
+  await stop(); // { shipped, left, errors, stopped }
+  process.exit(0);
+});
+```
+
+A program that ends by itself with `SIGNAFAIDE_HEAD_FILE` set writes the stop note at exit, within the exit limit above. `SIGNAFAIDE_INSTANCE_STOP=0` turns that off; `=1` turns it on without a head file.
+
+## Errors
+
+Service errors are `SignafaideError`. Its `message` is the service's one-sentence fix, word for word, and `code` is the error code. A `chain_conflict` is reported with the service's head and blocks shipping for that agent. The local chain is never rewritten.
+
+## Verify
+
+```bash
+signafaide verify                     # the local .signafaide state
+signafaide verify pack.json           # a pack from `signafaide export`
+signafaide verify pack.json --keys keys.json --json
+```
+
+There is one line per check, then the result line: `PASS` or `FAIL` and how many checks passed, are pending and were not checked, for example `PASS: 11 of 12 checks passed, 1 pending (public timestamp)`. Pending means co-signed, waiting for the public timestamp (an anchor is made every hour and usually confirms within a few hours); pending and not checked are never counted as passed. The exit code is 0 for PASS, 1 for FAIL and 2 when an input can't be read. Locally, a chain the witness refuses for good (or that an older SDK wrote with time going back) shows as `BLOCKED` with the reason and the next step, and always exits 1; `--fail-on-pending-send` also fails while records wait to be co-signed, for CI. Locally, before anything was co-signed, the witness's checks show `NOT CHECKED` ("nothing co-signed yet") and no request is made. Service keys come from `--keys <file|url>`, or by default from the service's `/.well-known/signafaide-keys.json`. `--trust-embedded-keys` uses the keys inside the pack and prints a warning. The drand BLS signature is checked too (through `@noble/curves`, a dependency).
+
+```ts
+import { verify } from "signafaide";
+const report = await verify("pack.json", { keys: "keys.json" });
+console.log(String(report)); // the same lines
+report.ok;
+```
+
+### Browsers and edge runtimes
+
+`signafaide/core` has no Node APIs. It provides JCS, hashing, Ed25519 keys and signatures, Merkle proofs, the HTTP client and `verifyPack`:
+
+```ts
+import { verifyPack, loadDrandVerifier } from "signafaide/core";
+const report = verifyPack(pack, serviceKeys, { drand: await loadDrandVerifier() });
+```
+
+`record`, `wrapTools` and the CLI need Node 18+ because they keep local state on disk.
+
+
+**For an auditor** the agent keys must come from you, not from the pack (otherwise a service could add a key of its own): `signafaide keys export -o agent-keys.json` writes every agent key, each signed by itself. The auditor runs `npx signafaide verify pack.json --keys <witness>/.well-known/signafaide-keys.json --agent-keys agent-keys.json --head head.json` (`head.json` is your `.signafaide/head.json`, which catches records cut off at the end). Without `--agent-keys` the client-signature check says `NOT INDEPENDENT` and the result is FAIL; `--trust-embedded-agent-keys` accepts the pack's keys and says so in the result line. On your own machine, `signafaide verify pack.json` uses this project's agent keys automatically. Public timestamps are checked online (`--timestamp-check <url>`, default `https://mempool.space/api`); with `--offline` they show `NOT CHECKED`, never confirmed. `signafaide keys rotate --agent X` gives an agent a new key; the SDK signs the registration of the new key with the old one.
+
+## Replay
+
+```bash
+npx signafaide replay --agent claims-bot --from 2026-11-01 --to 2026-11-13 --html replay.html
+npx signafaide replay --agent claims-bot --record 3fa2b1c9
+```
+
+Shows the agent's decisions and outcomes in order, with their inputs and outputs from the local chain. Each record is checked against its `content_hash`, its signature and the service's receipt first, so an edited line shows as `FAIL`. `--html` also writes a self-contained page. In code: `const r = await replay({ agent: "claims-bot", since: "2026-11-01" })`, then `String(r)`, `r.toHtml()`, `r.ok`.
+
+## Experiments: A/B on past decisions
+
+```ts
+import { compare } from "signafaide";
+
+const report = await compare(async (input) => ({ approved: input.amount < 300 }), { agent: "claims-bot", name: "approve_payout", since: "2026-11-01", limit: 500 });
+console.log(String(report)); // n, agreed, changed, errors, then the changed decisions
+await report.upload();       // optional: hash-only unless content is stored with the service
+```
+
+The chain is verified first; then each recorded decision with a co-signed outcome is re-run with your candidate, on your machine, with your own models and keys, and compared with the recorded `output` (exact canonical-JSON equality, or `same: (recorded, candidate) => ...`). The candidate gets the input as recorded: the single argument, or the argument list. A candidate that throws counts as an error. **Candidates must be side-effect free:** compare never writes records, and a recorded tool called from a candidate throws instead of running. CLI: `npx signafaide compare --module ./candidate.mjs:approve --name approve_payout --upload`.
+
+## CLI
+
+| command | does |
+|---|---|
+| `signafaide init` | the opt-in: creates a free sandbox if there is no key (no account, email or card), prints the claim link |
+| `signafaide status` | key and env, agents, last receipt, outbox size, quota; BLOCKED chains (exit 1) |
+| `signafaide repair [--agent X] [--dry-run]` | sets aside records no witness can take (kept in `.signafaide/held/`, still verified), gives the agent a new key and starts a new segment |
+| `signafaide verify [pack.json] [--keys ...] [--agent-keys ...] [--head ...] [--offline] [--fail-on-pending-send] [--json]` | verifies; exit 0 PASS, 1 FAIL or BLOCKED, 2 unreadable input |
+| `signafaide keys export [-o agent-keys.json]` / `signafaide keys rotate [--agent X] [--env test\|live]` | the agent-key file for auditors / a new key for an agent |
+| `signafaide keys create [--agent X] [--env test\|live] [--print]` | a new key, registered with the witness, for a secret store; writes no record (`--print`: prints the key file and saves nothing here) |
+| `signafaide export [--agent X] [--from T] [--to T] [-o pack.json]` | downloads a verification pack of the whole project (every agent); `--agent` narrows it to one |
+| `signafaide replay [--agent X] [--from T] [--to T] [--record H] [--html out.html] [--json]` | the verified timeline from the local chain; exit 1 on FAIL |
+| `signafaide compare --module file.mjs:fn [--agent X] [--name N] [--from T] [--to T] [--limit N] [--upload [--values]]` | re-runs recorded decisions with a candidate and compares; `--values` sends the values too (stored-content decisions only) |
+| `signafaide claim` | prints the claim link again |
+| `signafaide flush` | ships the outbox now (with no key it sends nothing and says so) |
+| `signafaide agent setup [--agent <name>]... [--url <witness>] [--dry-run] [--dir <project>]` | installs the MCP server and the skill in your coding agents; `--url` sets `SIGNAFAIDE_URL` in each MCP entry |
+| `signafaide mcp` | runs the MCP server over stdio, when the separate `signafaide-mcp` package is installed (`agent setup` configures `npx -y signafaide-mcp@<version>` instead) |
+
+## Coding agents
+
+```bash
+npx signafaide agent setup
+```
+
+finds the coding agents installed or in use and, for each one, adds the signafAIde MCP server (`npx -y signafaide-mcp@<this version>`, a separate package so the SDK's own install stays small) and a short skill that says when signafAIde fits, in that agent's own format. It merges into existing files, never overwrites what is there, and running it twice changes nothing. `--agent` picks agents (`claude-code`, `codex`, `cursor`, `gemini-cli`, `copilot`, `windsurf` or `all`); `--dry-run` shows what would change; `--url <witness url>` writes `SIGNAFAIDE_URL` into each agent's MCP entry (a self-hosted witness). It also adds `.signafaide/` to `.gitignore` and remembers which coding agent set the project up. It ends with what is sent (nothing until a person opts in, then hashes only), where the content stays, and a reminder to give agents explicit names.
+
+| agent | MCP server | instructions |
+|---|---|---|
+| Claude Code | `.mcp.json` | `.claude/skills/signafaide/SKILL.md` and a `CLAUDE.md` section |
+| Codex | `$CODEX_HOME/config.toml` (default `~/.codex/config.toml`, shared by all projects) | `AGENTS.md` section |
+| Cursor | `.cursor/mcp.json` | `.cursor/rules/signafaide.mdc` |
+| Gemini CLI | `.gemini/settings.json` | `GEMINI.md` section |
+| GitHub Copilot (VS Code) | `.vscode/mcp.json` | `.github/copilot-instructions.md` section |
+| Windsurf | `.devin/mcp_config.json` | `AGENTS.md` section |
+
+The MCP tools: `signafaide_setup` (a preview that changes, creates and sends nothing: the language and framework, the install command, the code edit and what a sandbox would send), `signafaide_init` (creates the external sandbox, and nothing unless the agent passes `confirm_user_agreed: true` and `user_words`, what the person wrote when they agreed), `signafaide_quickstart`, `signafaide_verify`, `signafaide_status`, `signafaide_explain_error` and `signafaide_claim_link`. The MCP server needs Node.js 20 or later.
+
+## What is sent besides records
+
+Every request carries one header:
+
+```
+X-Signafaide-Client: signafaide-js/0.2.1 (node 22.22.0; adapter=vercel-ai; agent=cursor; via=sdk)
+```
+
+It holds the SDK and its version, the JavaScript runtime and its version, and the framework adapter in use (`langchain`, `openai-agents`, `vercel-ai` or `none`). It also holds the coding agent that set the project up (`claude-code`, `codex`, `cursor`, `gemini-cli`, `copilot`, `windsurf` or `unknown`) and how the call was made (`sdk`, `cli` or `mcp`). The agent is found from that tool's environment variables when the project is set up (`signafaide init`, `signafaide agent setup`) and kept in `.signafaide/config.json`; it is never guessed per process, so a program started from a terminal inside a coding agent is not labelled as that agent. `signafaide verify` and `signafaide export` also send `POST /v1/events {"name": "verify"}` (or `"export"`) when there is an API key. If that fails, the command is not affected. With no API key nothing at all is sent.
+
+The header and the events hold no content, file paths, hostnames or user names. `SIGNAFAIDE_TELEMETRY=0` removes the adapter, the coding agent and the strict-mode flag from the header, stops the coding agent being saved in `.signafaide/config.json`, and turns the events off. The SDK and its version, the runtime and its version, and how the call was made (`via`) are still sent.
+
+## Configuration
+
+| variable | default |
+|---|---|
+| `SIGNAFAIDE_API_KEY` | read from the environment, then `.env`; created by `signafaide init`; none: local only |
+| `SIGNAFAIDE_URL` | `https://api.signafaide.co.uk`; a self-hosted witness, or `http://127.0.0.1:7474` for a local dev server. Alone it sends nothing |
+| `SIGNAFAIDE_HOME` | `.signafaide/` in the project root; a mounted volume or `/tmp/signafaide` in production |
+| `SIGNAFAIDE_AGENT_KEY_<NAME>` | an agent's key file content from a secret store (or `keyProvider` in code); live keys can use `SIGNAFAIDE_AGENT_KEY_LIVE_<NAME>`, made with `signafaide keys create --env live --print` |
+| `SIGNAFAIDE_HEAD_FILE` | unset: a file outside `.signafaide/` (a persistent volume) that also keeps the newest signed heads, for containers whose disk does not last; or `onHead` in `configure()` |
+| `SIGNAFAIDE_INSTANCE_STOP` | unset: a signed stop note is written at a clean exit when a head file is set; `1` always, `0` never (see Containers that restart) |
+| `SIGNAFAIDE_EXIT_FLUSH_MS` | `3000`: the longest a finished program is kept open to send; `0` never |
+| `SIGNAFAIDE_TEST` | test mode is on by itself under Vitest, Jest and `node --test` (records in a temporary folder, nothing sent); `1` forces it, `0` turns it off |
+| `SIGNAFAIDE_DISABLED` | `1`: wrapped functions run and nothing is recorded |
+| `SIGNAFAIDE_SEGMENT_BYTES` | `67108864`: a shipped chain file larger than this rolls into a sealed segment |
+| `SIGNAFAIDE_TELEMETRY` | `0` turns off the adapter, agent and strict fields and the events |
+
+Running in containers, with replicas, on read-only or ephemeral disks, on serverless or in CI: see the [production guide](https://signafaide.co.uk/docs/production). A record's time never goes back along its chain: if the clock steps back, the record keeps the previous record's time and the clock's reading goes in `meta.wall_time`.
+
+Each project and env has its own chain per agent. A new key, for example a live key replacing the sandbox key, starts a fresh chain at seq 1. The old chain stays on disk, and `signafaide status` lists every chain.
+
+The project root is the nearest folder with `.signafaide/`, `.git`, `pyproject.toml` or `package.json`. `signafaide init` and `agent setup` add `.signafaide/` to `.gitignore` (never a running program). Agent keys live in `.signafaide/agents/<agent>.key` (sandbox) and `.signafaide/agents/live/<agent>.key` (live), mode 0600, and never leave the machine. Back them up. The Python and TypeScript SDKs share this folder and its formats.
+
+ESM and CommonJS, with types. Dependencies: `@noble/ed25519`, `@noble/hashes` and `@noble/curves` (drand). The MCP server is the separate `signafaide-mcp` package, with the MCP SDK and `zod`, so `npm install signafaide` stays small.