psyclaw@0.27.4
A standalone, evidence-grounded social-science research agent — a conversation-driven agent that bundles the Pi runtime with psyclaw's research contracts, evidence gates, and workflows
AI Security Review
scanned 2mo ago · by lpm-firewall-aiSource inspection found a user-invoked research-agent CLI with guarded configuration and subprocess features, not malicious install-time behavior. The static hints reflect legitimate capability primitives rather than a concrete attack chain.
Static reason
No blocking static signals were detected.; source matched previously finalized malicious package; routed for review; previous stored version diff introduced dangerous source
Trigger
A user runs psyclaw commands or invokes its Pi extension.
Impact
It can configure selected Pi providers or run configured local tools, but no unconsented install-time mutation or data exfiltration was found.
Mechanism
User-invoked research-agent configuration and local tool integration.
AI rationale
Source inspection found a user-invoked research-agent CLI with guarded configuration and subprocess features, not malicious install-time behavior. The static hints reflect legitimate capability primitives rather than a concrete attack chain.
Evidence
package.jsondist/src/cli.jsdist/src/setup.jsdist/src/integrations/mcp.jsdist/src/install/installer.jsdist/src/agents/discover.jsdist/src/updates/update.js
Decision evidence
public snapshotAI called this Clean at 94.0% confidence as Benign with low false-positive risk.
Evidence for AI clean decision
- An explicit setup command can write selected provider metadata to the Pi agent directory.
- The package can spawn configured MCP child processes, but it uses argument arrays, disables the shell, and passes an explicit environment.
Evidence against
- package.json has only a prepublish-only hook, so registry installation does not run package code.
- There is no runtime dependency on psyclaw itself.
- Agent discovery only checks paths and does not read credential contents or execute installs.
- The agent installer rejects agent installs until a verified staged-artifact path exists.
- No source evidence shows credential exfiltration, hidden persistence, destructive behavior, or an unconsented install-time control-surface mutation.
Behavioral surface
ChildProcessCryptoDynamicRequireEnvironmentVarsFilesystemNetworkShell
HighEntropyStrings
Source & flagged code
4 flaggeddist/src/setup.jsView file
•matchType = previous_version_dangerous_delta
matchedPackage = psyclaw@0.26.2
matchedIdentity = npm:cHN5Y2xhdw:0.26.2
similarity = 0.850
summary = stored previous version shares package body but lacks this dangerous source file
High
Previous Version Dangerous Delta
This package version adds a dangerous source file absent from the previous stored version; route for source-aware review.
dist/src/setup.jsView on unpkg172const entry = fileURLToPath(import.meta.resolve("@earendil-works/pi-coding-agent"));
L173: const { readStoredCredential } = await import(pathToFileURL(join(dirname(entry), "core", "auth-storage.js")).href);
L174: return readStoredCredential(providerId, join(agentDir, "auth.json")) !== undefined;
Medium
Dynamic Require
Package source references dynamic require/import behavior.
dist/src/setup.jsView on unpkg · L172345export async function setupProviders(options = {}) {
const agentDir = options.agentDir ?? getAgentDir();
const modelsPath = join(agentDir, "models.json");
const selected = options.providers && options.providers.length > 0
? options.providers
: PROVIDER_PRESETS.map((preset) => preset.id);
Low
Findings
2 High4 Medium6 Low
HighKnown Malware Source Similaritydist/src/integrations/mcp.js
HighPrevious Version Dangerous Deltadist/src/setup.js
MediumDynamic Requiredist/src/setup.js
MediumNetwork
MediumEnvironment Vars
MediumStructural Risk Force Deep Review
LowNon Install Lifecycle Scripts
LowScripts Present
LowFilesystem
LowHigh Entropy Strings
LowUrl Strings
LowAi Review Evidencedist/src/setup.js
Affected versions and remediation
This report applies to psyclaw@0.27.4.
See all versions of psyclaw for other recorded verdicts.
- Review the evidence and your use of psyclaw@0.27.4 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: .
Think this verdict is wrong? Dispute it on GitHub or read how reports are produced.