rpi

Warn

Audited by Runlayer on Mar 15, 2026

Risk Level: MEDIUM
Scan Summary
Max Score
84%
Files
12
Flagged
8
Chunks
13
Flagged Files (8)
rpi/SKILL.mdMEDIUM
84.2%

Prompt Injection

Resource Abuse

Shadow Persistence

Guardrail Bypass

rpi/references/phase-data-contracts.mdMEDIUM
83.0%

Risky tool definition detected

| Transition | Output | Extraction | Input to Next | |------------|--------|------------|---------------| | → Discovery | Goal string + repo execution profile contract | Goal from the `/rpi` invocation; repo policy from `docs/contracts/repo-execution-profile.md` and `repo-execution-profile.schema.json` | `repo_profile` state is loaded before research/planning begins | | Discovery → Implementation | Epic execution context + discovery summary + `execution_packet` | `phased-state.json` + `.agents/r

rpi/references/context-windowing.mdMEDIUM
77.5%

Malicious tool definition detected

Continue until all shards are `done`. This keeps CPU and context budgets bounded while preserving full-file coverage.

rpi/references/error-handling.mdLOW
72.9%

Malicious tool definition detected

rpi/references/gate4-loop-and-spawn.mdLOW
71.7%

Resource Abuse

Shadow Persistence

rpi/references/complexity-scaling.mdLOW
70.7%

Malicious tool definition detected

rpi/references/troubleshooting.mdLOW
56.4%

Tool passed security scan

rpi/references/examples.mdLOW
56.3%

Tool passed security scan

Passed Files (4)Click to expand
rpi/references/gate-retry-logic.mdOK
35.4%

Tool passed security scan

rpi/references/report-template.mdOK
33.3%

Tool passed security scan

rpi/scripts/validate.shOK
26.3%

Tool passed security scan

rpi/references/phase-budgets.mdOK
15.9%

Tool passed security scan

Audit Metadata
Max File Score
84%
Classification
KNOWN_SERVER_KNOWN_TOOL_LIST
Files Scanned
12
Files Flagged
8
Chunks Analyzed
13
Analyzed
Mar 15, 2026, 01:54 PM
Security Audit — runlayer — rpi