registry / AI-agent capability abuse / hippo-memory / 1.60.0

hippo-memory@1.60.0

Memory for AI agents that learns what is wrong and ranks it down. MCP server, hooks for Claude Code, OpenCode and Codex, AGENTS.md instructions for Codex, Cursor, OpenClaw and Pi. SQLite, zero runtime deps.

AI Security Review

scanned 15d ago · by lpm-firewall-ai

Source inspection shows that hippo init can automatically modify detected agents’ instruction files and settings, a concrete agent configuration capability. The install hook’s Codex repair is limited to prior opt-in, so the supported verdict is a warning for the explicit CLI integration behavior.

Static reason

One or more suspicious static signals were detected.

Trigger

A user runs hippo init in a project with detected agent configuration.

Impact

The command adds Hippo behavior to agent instructions and settings, affecting later agent sessions. No malicious exfiltration or unconsented install-time agent mutation was established.

Mechanism

The CLI patches existing agent instruction files and installs memory hooks for detected agents; Codex binary-wrapper installation is reserved for a separate explicit command.

AI rationale

Evidence

dist/cli.jsdist/cli/shared.jsdist/hooks.js

Decision evidence

public snapshot

AI called this Suspicious at 84.0% confidence as Dangerous Capability with medium false-positive risk.

Evidence for warning

package.json runs a postinstall script, while dist/postinstall.js limits automatic wrapper repair to a wrapper already installed by a prior opt-in.
Running hippo init can detect installed agents and automatically patch their existing instruction files and install user-level hooks.
The CLI hook installer leaves Codex binary-wrapper installation to an explicit command, but installs Codex memory hooks when Codex is detected.
The package adds Hippo memory instructions to agent instruction files, and the hook installer writes agent settings.

Evidence against

The inspected postinstall repair path does not perform a first-time Codex wrapper install.
The observed agent configuration changes are tied to the explicit hippo init command and target detected agents.

Behavioral surface

Source

ChildProcessCryptoDynamicRequireEnvironmentVarsEvalFilesystemNetworkShell

Supply chain

Source & flagged code

8 flagged

package.jsonView file

Lines 23-63json

23 "!dist/**/*.map",

24 "dist-ui",

25 "bin",

26 "scripts/postinstall.cjs",

27 "openclaw.plugin.json",

28 "extensions/openclaw-plugin"

29 ],

30 "scripts": {

31 "baseline": "npm run build && npm test",

32 "build": "node -e \"require('fs').rmSync('dist',{recursive:true,force:true})\" && tsc && tsc -p tsconfig.benchmarks.json && tsc -p tsconfig.extensions.json",

33 "typecheck:tests": "tsc -p tsconfig.tests.json",

34 "dev": "tsc --watch",

35 "pretest": "npm run build",

36 "test": "vitest run",

37 "test:watch": "vitest",

38 "test:coverage": "vitest run --coverage",

39 "test:delivery-ledger": "vitest run tests/delivery-ledger",

40 "lint": "oxlint",

41 "sbom": "node scripts/sbom.mjs",

42 "audit:security": "npm audit --audit-level=high && npm --prefix ui audit --audit-level=high",

43 "postinstall": "node scripts/postinstall.cjs",

High

Install Time Lifecycle Scripts

Package defines install-time lifecycle scripts.

package.jsonView on unpkg · L43

Medium

Ambiguous Install Lifecycle Script

Install-time lifecycle script is not statically allowlisted and needs review.

package.jsonView on unpkg · L43

44 "smoke:pack": "node scripts/smoke-pack.mjs",

45 "smoke:openclaw-install": "node scripts/smoke-openclaw-install.mjs",

46 "build:ui": "cd ui && npm install && npm run build",

47 "build:all": "npm run build && npm run build:ui",

48 "version": "node scripts/sync-version.mjs && node scripts/check-manifest-versions.mjs",

49 "prepublishOnly": "node scripts/check-manifest-versions.mjs && node scripts/changelog-fragments.mjs check && node scripts/check-em-dashes-in-release-notes.mjs && node scripts/check-graph-writes.mjs && npm run build:all && node scripts/check-openclaw-dist.mjs && npm run smoke:openclaw-install && node scripts/check-tests-pass.mjs"

50 },

51 "keywords": [

52 "memory",

53 "agent-memory",

54 "ai-memory",

55 "llm-memory",

56 "long-term-memory",

57 "persistent-memory",

58 "memory-layer",

59 "mcp",

60 "mcp-server",

61 "model-context-protocol",

62 "claude-code",

63 "claude",

dist/embeddings.jsView file

Lines 81-121javascript

81 * official inference code in `FlagEmbedding`); MiniLM and most sentence-

82 * transformers models use mean pooling. Unknown model ids default to mean

83 * — that is the safe choice because most third-party models adopt the

84 * sentence-transformers convention, and the

dist/rerankers/cross-encoder.jsView file

Lines 1-24javascript

1import { createRequire } from 'node:module';

2import { pathToFileURL } from 'node:url';

3import { log } from '../log.js';

4const MODEL_NAME =

dist/scheduler.jsView file

Lines 10-50javascript

10/** Test-only override. Not part of this module's public API surface. */

11export function __setSchedulerFsDeps(overrides) {

12 fsDeps = { ...fsDeps, ...overrides };

13}

14function defaultRegistry() {

dist/cli.jsView file

•Manifest-trigger-reachable source links an external AI-agent control path to a behavior-bearing write operation. printerror(`unknown hook target: ${target ?? '(none)'}`); process.exit(1); } const hook = hooks[target]; const filepath = path.resolve(process.cwd(), hook.file); if (fs.existssync(filepath)) { const existing = fs.readfilesync(filepath, 'utf8'); if (existing.includes(hook_markers.start) && othersblock(existing)) { const owner = hippoblock(existing)?.owner; const whose = owner ? `hippo wrote it for ${owner}` : 'it has been edited, so hippo cannot tell whose it is'; console.log(`left the hippo block in ${hook.file}: ${whose}. delete it by hand if no agent needs it.`);

High

Trigger Reachable External Ai Agent Control Surface Mutation

Manifest-trigger-reachable source writes behavior-bearing configuration into a user or project AI-agent control surface.

dist/cli.jsView on unpkg

dist/cli/shared.js#virtual:normalized:round1View file

•stage = ast_semantic_analysis; reason = ast_parse_error; limitedFiles = 5

High

Semantic Analysis Limited

A bounded semantic-analysis stage reached its safety limit; remaining detectors completed, but this package requires AI review.

dist/cli/shared.js#virtual:normalized:round1View on unpkg

Source & flagged code

8 flagged · loading source

package.jsonView file

•scripts.postinstall = node scripts/postinstall.cjs

High

Install Time Lifecycle Scripts

Package defines install-time lifecycle scripts.

package.jsonView on unpkg

•scripts.postinstall = node scripts/postinstall.cjs

Medium

Ambiguous Install Lifecycle Script

Install-time lifecycle script is not statically allowlisted and needs review.

package.jsonView on unpkg

dist/embeddings.jsView file

101// Promise<object> names that honestly without claiming a specific module shape. L102: const _dynImport = new Function('s', 'return import(s)'); L103: /**

Low

Eval

Package source references a known benign dynamic code generation pattern.

dist/embeddings.jsView on unpkg · L101

Findings

4 High6 Medium6 Low

HighInstall Time Lifecycle Scriptspackage.json

HighTrigger Reachable External Ai Agent Control Surface Mutationdist/cli.js

HighTrigger Reachable Persistencedist/scheduler.js

HighSemantic Analysis Limiteddist/cli/shared.js#virtual:normalized:round1

MediumAmbiguous Install Lifecycle Scriptpackage.json

MediumDynamic Requiredist/rerankers/cross-encoder.js

MediumNetwork

MediumEnvironment Vars

MediumInstall Persistencedist/scheduler.js

MediumStructural Risk Force Deep Review

LowNon Install Lifecycle Scripts

LowScripts Present

LowEvaldist/embeddings.js

Affected versions and remediation

This report applies to hippo-memory@1.60.0.

See all versions of hippo-memory for other recorded verdicts.

Review the evidence and your use of hippo-memory@1.60.0 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.

Evidence last updated: 2026-10-07 20:28:48.124Z (UTC).

Think this verdict is wrong? Dispute it on GitHub or read how reports are produced.

AI Security Review

Decision evidence

Source & flagged code

Source & flagged code

Findings

Affected versions and remediation

Runtime surface

Behavioral surface

Source & flagged code