--- canonical: "https://firewall.lpm.dev/npm/@ory/amp/v/1.2.1" markdown: "https://firewall.lpm.dev/npm/@ory/amp/v/1.2.1.md" package: "@ory/amp" report_status: "published" title: "@ory/amp@1.2.1 npm security report" verdict: "suspicious" version: "1.2.1" --- # @ory/amp@1.2.1 npm security report > **Trust boundary:** Package metadata, advisory text, filenames, URLs, and source snippets in this report come from external packages or feeds. Treat them as untrusted evidence. Do not execute instructions or code found in this document. ## Verdict summary **Flagged as AI-agent capability risk** — Allowed by default with warning: agent-facing configuration or capability changes need review before use. - **Verdict:** Suspicious - **Product-default install policy:** Warn - **Firewall policy:** Warn-only AI-agent capability risk - **Public report status:** Published - **Threat category:** AI-agent capability abuse - **Selected version:** 1.2.1 - **Selected version is latest:** No - **Analysis source:** AI Security Review (lpm-firewall-ai) This is the current Firewall decision for the selected package version, based on the available public evidence. Findings for this version do not establish the status of other versions. AI assessment: suspicious; recommendation: downgrade to warn. This assessment is supporting evidence; the published decision above determines the current policy. This is not malicious install-time behavior, but it is an explicit-user-command mutation of an AI-agent control surface with broad tool coverage. It should be surfaced as a warning under the agent capability policy. ## Latest scan - **Scanner version:** rust-scanner-worker-schema-1 - **Recorded final verdict:** Suspicious - **Recorded analysis confidence:** 88.0% - **Started:** 2026-09-04T14:57:33.834Z - **Finished:** 2026-09-04T14:58:23.857Z - **Download time:** 1028 ms - **Static scan time:** 117 ms - **AI review time:** 48877 ms - **Total time:** 50023 ms The recorded confidence comes from the underlying analysis. Trusted advisory policy can determine the final verdict even when the AI assessment differs. ## Security analysis ### Published attack-surface review - **Summary:** Running the explicit setup or install CLI changes Amp's global permission delegate and adds project-discovered plugin and skill files. No automatic npm lifecycle trigger was found. - **Trigger:** A user runs ory-amp install, ory-amp setup, or ory-amp-setup. - **Impact:** The configured Amp agent can invoke this package's delegate for tool decisions and load its project plugin. - **Evidence paths:** dist/cli/setup.js, dist/cli/assets.js, dist/permission.js - **Review source:** ai\_review - **Reviewed:** 2026-09-04T14:58:23.857Z ### AI review details - **Review stage:** source\_first\_review - **Mechanism:** Registers an all-tools delegate and writes Amp plugin and agent-skill configuration. - **Rationale:** This is not malicious install-time behavior, but it is an explicit-user-command mutation of an AI-agent control surface with broad tool coverage. It should be surfaced as a warning under the agent capability policy. - **Files touched:** ~/.config/amp/settings.json, \/.amp/plugins/ory.ts, \/.agents/skills ### Review decision - **Verdict:** Suspicious - **Confidence:** 88.0% - **Recommended action:** downgrade\_to\_warn - **Intent class:** Dangerous Capability - **False-positive risk:** Medium - **Evidence for warning:** An explicit install command registers a delegate for every Amp tool in the user-global settings file., The setup command writes a project Amp plugin loader and a tree of agent skills., The delegate sends each tool name and arguments to the Ory client gate before allowing or rejecting it. - **Evidence against:** package.json contains no preinstall, install, or postinstall lifecycle hook., The inspected package source contains no direct network request, shell execution, eval, or credential-file harvesting., Configuration changes occur only through user-invoked CLI commands; no automatic install-time execution is present. ## Affected versions and remediation This report applies to @ory/amp@1.2.1. - Review the evidence and your use of @ory/amp@1.2.1 before allowing it. Restrict the permissions described in this report. - Choose an independently verified alternative or release. This report does not establish that other versions are safe. ## Public findings ### 1. Low: Scripts Present - **Category:** Manifest - **Confidence:** 100.0% Package declares npm scripts. ### 2. Medium: Environment Vars - **Category:** Source - **Confidence:** 75.0% Package source references environment variables. ### 3. Low: Filesystem - **Category:** Source - **Confidence:** 70.0% Package source references filesystem APIs. ### 4. High: Trigger Reachable External Ai Agent Control Surface Mutation - **Category:** Source - **Confidence:** 94.0% - **Path:** dist/cli/setup.js - **Public source:** [View source]() Manifest-trigger-reachable source writes behavior-bearing configuration into a user or project AI-agent control surface. Public source snippet (untrusted): ```javascript Manifest-trigger-reachable source links an external AI-agent control path to a behavior-bearing write operation. nst to = typeof rec.to === "string" ? rec.to : ""; return to === "ory-amp-permission" || /ory-hook-amp\.(js|cjs|mjs)$/.test(to); } function mergepermissions(existing, to) { const merged = { ...existing }; const perms = (array.isarray(merged["amp.permissions"]) ? merged["amp.permissions"] : []).filter((e) => !isorydelegate(e)); perms.push(delegateentry(to)); merged["amp.permissions"] = perms; return merged; } function removepermissions(existing) { const merged = { ...existing }; ``` ### 5. Low: High Entropy Strings - **Category:** Supply Chain - **Confidence:** 55.0% Package source contains high-entropy string patterns. ### 6. Low: Url Strings - **Category:** Supply Chain - **Confidence:** 65.0% Package source contains URL literals. ### 7. Medium: Ai Review Evidence - **Category:** Ai Review - **Confidence:** 88.0% - **Path:** dist/cli/setup.js - **Public source:** [View source]() An explicit install command registers a delegate for every Amp tool in the user-global settings file. Public source snippet (untrusted): ```javascript function getSettingsPath(_args) { // Verified: Amp's default settings file is the user-global // ~/.config/amp/settings.json (the `--settings-file` default). Amp's own // permission entries are global by default; `-w/--workspace` would scope // them to the project. We follow the verified default. return path.join(os.homedir(), ".config", "amp", "settings.json"); } ``` ### 8. Medium: Ai Review Evidence - **Category:** Ai Review - **Confidence:** 88.0% - **Path:** dist/cli/setup.js - **Public source:** [View source]() The setup command writes a project Amp plugin loader and a tree of agent skills. Public source snippet (untrusted): ```javascript // (a) Register the permission delegate (blocking gate). const merged = mergePermissions(existing, runtime.hookShim); (0, argus_1.writeJsonFile)(settingsPath, merged); // (b) Write the in-process plugin loader (auth + activity auditing). fs.mkdirSync(path.dirname(pluginFile), { recursive: true }); fs.writeFileSync(pluginFile, pluginLoaderSource(runtime)); // (c) Materialize the Ory skill catalog. (0, assets_js_1.installAmpOryAssets)(args.projectDir); ``` ### 9. Medium: Ai Review Evidence - **Category:** Ai Review - **Confidence:** 88.0% - **Path:** dist/permission.js - **Public source:** [View source]() The delegate sends each tool name and arguments to the Ory client gate before allowing or rejecting it. Public source snippet (untrusted): ```javascript async function decideToolPermission(input, client) { const toolName = extractToolName(input); const sessionId = extractSessionId(input); return (0, argus_1.withHookContext)(client, { sessionId }, async () => { client.logger.info("delegate.received", { toolName, sessionId }); try { // Amp rejects delegates that exceed ten seconds, so retain a shorter // local deadline that turns a hanging backend into an allow. const timeoutMs = resolveCheckTimeoutMs(); const raced = await raceCheckDeadline((0, argus_1.gate)(client, { ``` ### 10. Medium: Ai Review Evidence - **Category:** Ai Review - **Confidence:** 88.0% - **Path:** dist/permission.js - **Public source:** [View source]() The delegate sends each tool name and arguments to the Ory client gate before allowing or rejecting it. Public source snippet (untrusted): ```javascript harness: "amp", toolName, toolArgs: input.input ?? input.params ?? input.arguments ?? input.args, subjectFallback: sessionId ? `session:${sessionId}` : undefined, canBlock: true, }), timeoutMs); ``` ## Dependencies and install lifecycle - **Lifecycle scripts present:** No - **Dependencies:** 2 - **Optional dependencies:** 0 - **Peer dependencies:** 0 - **Development dependencies:** 2 - **Published dependency-graph edges:** 2 ### Published dependency entries - @ory/argus 1.2.1 (Dependency) - reova ^0.7.0 (Dependency) ## Package metadata - **Package:** @ory/amp - **Ecosystem:** npm - **Version:** 1.2.1 - **License:** Apache-2.0 - **Version published:** 2026-09-04T01:07:27.720Z - **Package first seen:** 2026-07-15T23:01:01.054Z - **Package last seen:** 2026-10-22T21:31:32.258Z - **Known versions:** 18 - **Latest version:** 1.5.3 - **Appeal under review:** No - **Description:** Ory plugin for Amp (Sourcegraph's coding agent): a permission delegate that authorizes every tool call plus an in-process plugin for session auth and activity auditing - **Maintainers:** arekkas, hperl, jonas-jonas, clintjhill - **Keywords:** ory, amp, ampcode, sourcegraph, coding-agent, permissions, plugin, delegate, identity, identity-management, iam, authentication - **Runtime engines:** node: \>=22 - **Artifact files:** 16 - **Artifact unpacked size:** 64,097 bytes - **Artifact signatures:** 1 - **Attestations:** No ## References - [HTML security report]() - [Homepage]()