ambitry 0.1.0

package/LICENSE

@@ -0,0 +1,202 @@
+                                 Apache License
+                           Version 2.0, January 2004
+                        http://www.apache.org/licenses/
+
+   TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
+
+   1. Definitions.
+
+      "License" shall mean the terms and conditions for use, reproduction,
+      and distribution as defined by Sections 1 through 9 of this document.
+
+      "Licensor" shall mean the copyright owner or entity authorized by
+      the copyright owner that is granting the License.
+
+      "Legal Entity" shall mean the union of the acting entity and all
+      other entities that control, are controlled by, or are under common
+      control with that entity. For the purposes of this definition,
+      "control" means (i) the power, direct or indirect, to cause the
+      direction or management of such entity, whether by contract or
+      otherwise, or (ii) ownership of fifty percent (50%) or more of the
+      outstanding shares, or (iii) beneficial ownership of such entity.
+
+      "You" (or "Your") shall mean an individual or Legal Entity
+      exercising permissions granted by this License.
+
+      "Source" form shall mean the preferred form for making modifications,
+      including but not limited to software source code, documentation
+      source, and configuration files.
+
+      "Object" form shall mean any form resulting from mechanical
+      transformation or translation of a Source form, including but
+      not limited to compiled object code, generated documentation,
+      and conversions to other media types.
+
+      "Work" shall mean the work of authorship, whether in Source or
+      Object form, made available under the License, as indicated by a
+      copyright notice that is included in or attached to the work
+      (an example is provided in the Appendix below).
+
+      "Derivative Works" shall mean any work, whether in Source or Object
+      form, that is based on (or derived from) the Work and for which the
+      editorial revisions, annotations, elaborations, or other modifications
+      represent, as a whole, an original work of authorship. For the purposes
+      of this License, Derivative Works shall not include works that remain
+      separable from, or merely link (or bind by name) to the interfaces of,
+      the Work and Derivative Works thereof.
+
+      "Contribution" shall mean any work of authorship, including
+      the original version of the Work and any modifications or additions
+      to that Work or Derivative Works thereof, that is intentionally
+      submitted to Licensor for inclusion in the Work by the copyright owner
+      or by an individual or Legal Entity authorized to submit on behalf of
+      the copyright owner. For the purposes of this definition, "submitted"
+      means any form of electronic, verbal, or written communication sent
+      to the Licensor or its representatives, including but not limited to
+      communication on electronic mailing lists, source code control systems,
+      and issue tracking systems that are managed by, or on behalf of, the
+      Licensor for the purpose of discussing and improving the Work, but
+      excluding communication that is conspicuously marked or otherwise
+      designated in writing by the copyright owner as "Not a Contribution."
+
+      "Contributor" shall mean Licensor and any individual or Legal Entity
+      on behalf of whom a Contribution has been received by Licensor and
+      subsequently incorporated within the Work.
+
+   2. Grant of Copyright License. Subject to the terms and conditions of
+      this License, each Contributor hereby grants to You a perpetual,
+      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+      copyright license to reproduce, prepare Derivative Works of,
+      publicly display, publicly perform, sublicense, and distribute the
+      Work and such Derivative Works in Source or Object form.
+
+   3. Grant of Patent License. Subject to the terms and conditions of
+      this License, each Contributor hereby grants to You a perpetual,
+      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+      (except as stated in this section) patent license to make, have made,
+      use, offer to sell, sell, import, and otherwise transfer the Work,
+      where such license applies only to those patent claims licensable
+      by such Contributor that are necessarily infringed by their
+      Contribution(s) alone or by combination of their Contribution(s)
+      with the Work to which such Contribution(s) was submitted. If You
+      institute patent litigation against any entity (including a
+      cross-claim or counterclaim in a lawsuit) alleging that the Work
+      or a Contribution incorporated within the Work constitutes direct
+      or contributory patent infringement, then any patent licenses
+      granted to You under this License for that Work shall terminate
+      as of the date such litigation is filed.
+
+   4. Redistribution. You may reproduce and distribute copies of the
+      Work or Derivative Works thereof in any medium, with or without
+      modifications, and in Source or Object form, provided that You
+      meet the following conditions:
+
+      (a) You must give any other recipients of the Work or
+          Derivative Works a copy of this License; and
+
+      (b) You must cause any modified files to carry prominent notices
+          stating that You changed the files; and
+
+      (c) You must retain, in the Source form of any Derivative Works
+          that You distribute, all copyright, patent, trademark, and
+          attribution notices from the Source form of the Work,
+          excluding those notices that do not pertain to any part of
+          the Derivative Works; and
+
+      (d) If the Work includes a "NOTICE" text file as part of its
+          distribution, then any Derivative Works that You distribute must
+          include a readable copy of the attribution notices contained
+          within such NOTICE file, excluding those notices that do not
+          pertain to any part of the Derivative Works, in at least one
+          of the following places: within a NOTICE text file distributed
+          as part of the Derivative Works; within the Source form or
+          documentation, if provided along with the Derivative Works; or,
+          within a display generated by the Derivative Works, if and
+          wherever such third-party notices normally appear. The contents
+          of the NOTICE file are for informational purposes only and
+          do not modify the License. You may add Your own attribution
+          notices within Derivative Works that You distribute, alongside
+          or as an addendum to the NOTICE text from the Work, provided
+          that such additional attribution notices cannot be construed
+          as modifying the License.
+
+      You may add Your own copyright statement to Your modifications and
+      may provide additional or different license terms and conditions
+      for use, reproduction, or distribution of Your modifications, or
+      for any such Derivative Works as a whole, provided Your use,
+      reproduction, and distribution of the Work otherwise complies with
+      the conditions stated in this License.
+
+   5. Submission of Contributions. Unless You explicitly state otherwise,
+      any Contribution intentionally submitted for inclusion in the Work
+      by You to the Licensor shall be under the terms and conditions of
+      this License, without any additional terms or conditions.
+      Notwithstanding the above, nothing herein shall supersede or modify
+      the terms of any separate license agreement you may have executed
+      with Licensor regarding such Contributions.
+
+   6. Trademarks. This License does not grant permission to use the trade
+      names, trademarks, service marks, or product names of the Licensor,
+      except as required for reasonable and customary use in describing the
+      origin of the Work and reproducing the content of the NOTICE file.
+
+   7. Disclaimer of Warranty. Unless required by applicable law or
+      agreed to in writing, Licensor provides the Work (and each
+      Contributor provides its Contributions) on an "AS IS" BASIS,
+      WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
+      implied, including, without limitation, any warranties or conditions
+      of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
+      PARTICULAR PURPOSE. You are solely responsible for determining the
+      appropriateness of using or redistributing the Work and assume any
+      risks associated with Your exercise of permissions under this License.
+
+   8. Limitation of Liability. In no event and under no legal theory,
+      whether in tort (including negligence), contract, or otherwise,
+      unless required by applicable law (such as deliberate and grossly
+      negligent acts) or agreed to in writing, shall any Contributor be
+      liable to You for damages, including any direct, indirect, special,
+      incidental, or consequential damages of any character arising as a
+      result of this License or out of the use or inability to use the
+      Work (including but not limited to damages for loss of goodwill,
+      work stoppage, computer failure or malfunction, or any and all
+      other commercial damages or losses), even if such Contributor
+      has been advised of the possibility of such damages.
+
+   9. Accepting Warranty or Additional Liability. While redistributing
+      the Work or Derivative Works thereof, You may choose to offer,
+      and charge a fee for, acceptance of support, warranty, indemnity,
+      or other liability obligations and/or rights consistent with this
+      License. However, in accepting such obligations, You may act only
+      on Your own behalf and on Your sole responsibility, not on behalf
+      of any other Contributor, and only if You agree to indemnify,
+      defend, and hold each Contributor harmless for any liability
+      incurred by, or claims asserted against, such Contributor by reason
+      of your accepting any such warranty or additional liability.
+
+   END OF TERMS AND CONDITIONS
+
+   APPENDIX: How to apply the Apache License to your work.
+
+      To apply the Apache License to your work, attach the following
+      boilerplate notice, with the fields enclosed by brackets "[]"
+      replaced with your own identifying information. (Don't include
+      the brackets!)  The text should be enclosed in the appropriate
+      comment syntax for the file format. We also recommend that a
+      file or class name and description of purpose be included on the
+      same "printed page" as the copyright notice for easier
+      identification within third-party archives.
+
+   Copyright 2026 Rushikesh Sonu
+
+   Licensed under the Apache License, Version 2.0 (the "License");
+   you may not use this file except in compliance with the License.
+   You may obtain a copy of the License at
+
+       http://www.apache.org/licenses/LICENSE-2.0
+
+   Unless required by applicable law or agreed to in writing, software
+   distributed under the License is distributed on an "AS IS" BASIS,
+   WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+   See the License for the specific language governing permissions and
+   limitations under the License.
+

package/README.md

@@ -0,0 +1,163 @@
+# ambitry
+
+**Audit log and permission layer for AI agents. Runs on your machine. Nothing is uploaded.**
+
+You cannot currently answer a simple question about your agents: *what did they do last Tuesday, and what were they able to do?* Ambitry is a proxy that sits in front of your LLM API and answers both — with a record of every call, a cap on what they can spend, and a switch that stops them.
+
+```bash
+npx ambitry
+```
+
+Then change one line in your agent:
+
+```python
+client = Anthropic(base_url="http://localhost:8787/anthropic")
+# or
+client = OpenAI(base_url="http://localhost:8787/openai/v1")
+```
+
+That's the integration. Open <http://localhost:8787> and you'll see every request, every tool your agent invoked, the arguments it passed, and what it cost.
+
+**No API key handy?**
+
+```bash
+npx ambitry --demo
+```
+
+Runs a scripted agent against a local stub — normal work, a leaked AWS key, then two actions stopped by policy. Nothing contacts a provider and no key is read from your environment.
+
+---
+
+## Controls
+
+An audit log tells you what happened. These stop it happening.
+
+**Spend caps.** Cut off agents that run away. Checked before the request is forwarded, so a runaway loop stops costing money immediately rather than after the fact.
+
+```json
+{ "spend": { "perHourUsd": 5, "perDayUsd": 50 } }
+```
+
+**Tool rules.** Deny wins over allow, and patterns must match the whole tool name.
+
+```json
+{ "tools": { "deny": ["delete_*", "send_email"] } }
+```
+
+**Deny by default**, once you know which tools you actually use:
+
+```json
+{ "default": "deny", "tools": { "allow": ["read_*", "search"] } }
+```
+
+**Kill switch.** One button in the dashboard halts every agent immediately.
+
+Put any of this in `ambitry.json` next to where you run the proxy. A malformed policy stops startup rather than falling back to permissive defaults — believing you have controls that never loaded is worse than having none.
+
+**Before you switch anything on, see what it would have done:**
+
+```bash
+$ npx ambitry dryrun
+
+  Policy dry run — 214 tool calls over 7 days
+
+  BLOCK  delete_user      3×  Blocked by deny rule "delete_*"
+  BLOCK  run_command     11×  Blocked by deny-by-default policy
+  allow  read_file      147×
+  allow  search_code     53×
+
+  14 calls that ran would have been stopped.
+```
+
+Nobody flips on deny-by-default blind, because the risk of silently breaking a working agent is too high. This answers the question from traffic you already have.
+
+---
+
+## Leak detection
+
+Ambitry scans every outgoing prompt for credentials and personal data before it reaches your provider — AWS keys, GitHub and Slack tokens, private keys, JWTs, payment cards, and high-entropy values assigned to credential-shaped names. If your agent has been quietly pasting a secret into its context, the dashboard says so on the front page.
+
+Detection is tuned against false positives, because a detector that fires on ordinary prose gets muted and then buys you false confidence. Payment card numbers are Luhn-checked so order IDs don't trip it, placeholder addresses like `user@example.com` are ignored, and a high-entropy string only counts if it's assigned to something that looks like a credential.
+
+**One thing to be clear about:** the finding itself keeps only a masked preview, but ambitry is an audit log — it stores your prompts. If a secret was in a prompt, it is in `ambitry.db`. Treat that file with the same care as the prompts themselves; it's plain SQLite on your disk, and you can delete it at any time.
+
+## What it sees, and what it doesn't
+
+Ambitry proxies the **LLM API**. That means it sees:
+
+- Every prompt, and every response
+- Every tool the model invoked, and the arguments it passed
+- Token counts and cost, per request
+- Which model, how long it took, whether it failed
+
+It does **not** see your agent's own outbound calls. If your code receives a tool call and then hits the Stripe or GitHub API directly, that request never passes through the proxy and is not recorded. Full network visibility needs an egress proxy with an installed CA certificate, which this is not.
+
+## How enforcement works, and its limits
+
+When a tool call violates policy, ambitry removes it from the response before your agent receives it. The agent never sees the call, so it cannot execute it, and the turn continues with a note explaining what was blocked.
+
+Two limits worth stating plainly:
+
+1. **It governs tools invoked through this proxy.** A tool your agent calls by some other path is outside its reach.
+2. **It is weaker than a framework-level hook.** Systems that intercept at the tool-execution boundary — a `before_tool_call` hook inside your agent framework — see every call regardless of where it came from. Ambitry intercepts one layer earlier, at the LLM response. That is real enforcement, and it is not the same thing.
+
+If you need guaranteed mediation of every action an agent takes, use a framework hook. If you want a record and a set of controls that work across any framework with a one-line change, this is that.
+
+---
+
+## Why this exists
+
+The alternatives are all either a cloud service or an enterprise platform:
+
+| | Where it runs | Enforcement | Cost of entry |
+|---|---|---|---|
+| Hosted LLM gateways | Their servers, holding your keys | Varies | Free tier, your prompts leave |
+| Self-hosted gateways | Docker + Postgres + ClickHouse + object storage | Varies | Real ops burden |
+| Enterprise agent-security platforms | Your infra, their control plane | Yes | Four to five figures a month |
+| **ambitry** | **Your machine** | **Yes** | **`npx ambitry`** |
+
+If your prompts can't go to a third party, and you're not standing up a ClickHouse cluster to read your own logs, there wasn't a good option. That's the gap.
+
+## Zero runtime dependencies
+
+`package.json` has no `dependencies` block. The proxy is `node:http` and `node:sqlite`; the tests are `node:test`; the dashboard is one HTML string with no external requests — no CDN, no fonts, no analytics.
+
+For a tool that reads every prompt you send, "there is no dependency tree to audit" is the point. Check it:
+
+```bash
+npm ls --omit=dev        # empty
+```
+
+Traces live in a SQLite file (`./ambitry.db`) that you own and can delete.
+
+---
+
+## Reference
+
+```
+npx ambitry [options]
+npx ambitry dryrun              what would this policy have blocked?
+npx ambitry models              list models with known pricing
+
+  --demo            scripted agent, no API key needed
+  --days <n>        window for dryrun, default 7
+  --port <n>        default 8787
+  --db <path>       default ./ambitry.db
+  --policy <path>   default ./ambitry.json
+```
+
+Point a provider at a different upstream — Azure, a gateway, a local model — with `AMBITRY_OPENAI_BASE` or `AMBITRY_ANTHROPIC_BASE`.
+
+Cost is computed from published per-token rates. Models with no known rate report `null` rather than a guess; a cost dashboard that quietly invents numbers is worse than one that admits ignorance.
+
+## Development
+
+```bash
+npm test        # 50 tests, no dependencies
+```
+
+The parser is verified boundary-independent: feeding a stream one byte at a time produces results identical to a single bulk read. That covers SSE events split across TCP chunks and multi-byte UTF-8 characters split mid-character — both of which corrupt audit data silently rather than crashing.
+
+## License
+
+Apache-2.0

package/dist/cli.d.ts

@@ -0,0 +1,2 @@
+#!/usr/bin/env node
+export {};

package/dist/cli.js

@@ -0,0 +1,114 @@
+#!/usr/bin/env node
+import { resolve } from 'node:path';
+import { Store } from "./db.js";
+import { Controls, loadPolicy } from "./policy.js";
+import { createServer } from "./server.js";
+import { pricedModels } from "./pricing.js";
+import { runDemo, DEMO_POLICY } from "./demo.js";
+import { dryRun, formatDryRun } from "./dryrun.js";
+const HELP = `ambitry — audit log and permission layer for AI agents
+
+  Runs on your machine. Nothing is uploaded.
+
+USAGE
+  npx ambitry [options]
+  npx ambitry dryrun          what would this policy have blocked?
+  npx ambitry models          list models with known pricing
+
+OPTIONS
+  --demo          run a scripted agent, no API key needed
+  --days <n>      window for dryrun                (default 7)
+  --port <n>      port to listen on            (default 8787)
+  --db <path>     trace database               (default ./ambitry.db)
+  --policy <path> policy file                  (default ./ambitry.json)
+  --help
+
+POINT YOUR SDK AT IT
+  Anthropic   base_url="http://localhost:8787/anthropic"
+  OpenAI      base_url="http://localhost:8787/openai/v1"
+
+POLICY (ambitry.json, all fields optional)
+  {
+    "default": "allow",
+    "tools": { "deny": ["delete_*", "send_email"] },
+    "spend": { "perHourUsd": 5, "perDayUsd": 50 }
+  }
+`;
+function flag(argv, name) {
+    const i = argv.indexOf(`--${name}`);
+    return i === -1 ? undefined : argv[i + 1];
+}
+function main(argv) {
+    if (argv.includes('--help') || argv.includes('-h')) {
+        process.stdout.write(HELP);
+        return;
+    }
+    if (argv[0] === 'models') {
+        process.stdout.write(pricedModels().join('\n') + '\n');
+        return;
+    }
+    if (argv[0] === 'dryrun') {
+        const days = Number(flag(argv, 'days') ?? 7);
+        const store = new Store(resolve(flag(argv, 'db') ?? 'ambitry.db'));
+        try {
+            const policy = loadPolicy(resolve(flag(argv, 'policy') ?? 'ambitry.json'));
+            process.stdout.write(formatDryRun(dryRun(store, policy, days), days));
+        }
+        catch (err) {
+            process.stderr.write(`\n  Policy error: ${err.message}\n\n`);
+            process.exitCode = 1;
+        }
+        finally {
+            store.close();
+        }
+        return;
+    }
+    const demo = argv.includes('--demo');
+    const port = Number(flag(argv, 'port') ?? 8787);
+    // Demo traffic goes to its own file so it never mixes with real traces.
+    const dbPath = resolve(flag(argv, 'db') ?? (demo ? 'ambitry-demo.db' : 'ambitry.db'));
+    const policyPath = resolve(flag(argv, 'policy') ?? 'ambitry.json');
+    let controls;
+    try {
+        controls = new Controls(demo ? DEMO_POLICY : loadPolicy(policyPath));
+    }
+    catch (err) {
+        // A policy that fails to load must stop startup. Falling back to the
+        // permissive default would leave you believing rules are in force that
+        // never loaded — the one failure mode a security tool must not have.
+        process.stderr.write(`\n  Policy error: ${err.message}\n\n`);
+        process.exit(1);
+    }
+    const store = new Store(dbPath);
+    const server = createServer({ store, controls });
+    let stopDemo;
+    server.listen(port, async () => {
+        if (demo) {
+            process.stdout.write(`\n  ambitry demo — enforcing policy\n`);
+            stopDemo = await runDemo(port, (s) => process.stdout.write(s));
+            return;
+        }
+        const mode = controls.enforcing ? 'enforcing policy' : 'logging only';
+        process.stdout.write(`\n  ambitry — ${mode}\n\n` +
+            `  Anthropic   base_url="http://localhost:${port}/anthropic"\n` +
+            `  OpenAI      base_url="http://localhost:${port}/openai/v1"\n\n` +
+            `  Dashboard   http://localhost:${port}\n` +
+            `  Traces      ${dbPath}\n\n`);
+    });
+    server.on('error', (err) => {
+        if (err.code === 'EADDRINUSE') {
+            process.stderr.write(`\n  Port ${port} is in use. Try: npx ambitry --port ${port + 1}\n\n`);
+            process.exit(1);
+        }
+        throw err;
+    });
+    const shutdown = () => {
+        stopDemo?.();
+        server.close();
+        store.close();
+        process.exit(0);
+    };
+    process.on('SIGINT', shutdown);
+    process.on('SIGTERM', shutdown);
+}
+main(process.argv.slice(2));

package/dist/db.d.ts

@@ -0,0 +1,94 @@
+export type Decision = 'allow' | 'deny';
+export interface TraceRow {
+    id: string;
+    started_at: number;
+    ended_at: number | null;
+    provider: string;
+    path: string;
+    model: string | null;
+    status: number | null;
+    streamed: number;
+    input_tokens: number | null;
+    output_tokens: number | null;
+    cost_usd: number | null;
+    request_json: string;
+    response_json: string | null;
+    error: string | null;
+}
+export interface ToolCallRow {
+    id: string;
+    trace_id: string;
+    seq: number;
+    name: string;
+    input_json: string;
+    decision: Decision;
+    reason: string | null;
+}
+export declare class Store {
+    private db;
+    constructor(path: string);
+    beginTrace(input: {
+        provider: string;
+        path: string;
+        model: string | null;
+        streamed: boolean;
+        request: unknown;
+    }): string;
+    finishTrace(id: string, patch: {
+        status?: number;
+        inputTokens?: number | null;
+        outputTokens?: number | null;
+        costUsd?: number | null;
+        response?: unknown;
+        error?: string | null;
+    }): void;
+    recordToolCall(t: {
+        traceId: string;
+        seq: number;
+        name: string;
+        input: unknown;
+        decision: Decision;
+        reason?: string | null;
+    }): void;
+    /**
+     * Persist what a scan found.
+     *
+     * Only the kind, severity, and masked preview are stored — never the secret
+     * itself. A tool that logs the credentials it catches has moved the leak,
+     * not closed it.
+     */
+    recordFindings(traceId: string, findings: {
+        kind: string;
+        severity: string;
+        preview: string;
+        count: number;
+    }[]): void;
+    findingsFor(traceId: string): {
+        kind: string;
+        severity: string;
+        preview: string;
+        count: number;
+    }[];
+    /** Aggregate across recent traffic, for the dashboard headline. */
+    findingsSummary(since: number): {
+        kind: string;
+        severity: string;
+        preview: string;
+        occurrences: number;
+        traces: number;
+    }[];
+    /** Total spend since a given epoch-ms timestamp. Used for spend caps. */
+    spendSince(since: number): number;
+    /** Tool calls in a window, for evaluating a policy against real history. */
+    toolCallsSince(since: number): {
+        name: string;
+        decision: Decision;
+        started_at: number;
+    }[];
+    recentTraces(limit?: number): TraceRow[];
+    trace(id: string): {
+        trace: TraceRow;
+        toolCalls: ToolCallRow[];
+    } | null;
+    close(): void;
+}