ctf-misc

Warn

Audited by Runlayer on Mar 16, 2026

Risk Level: MEDIUM
Scan Summary
Max Score
90%
Files
7
Flagged
5
Chunks
13
Flagged Files (5)
ctf-misc/SKILL.mdMEDIUM
89.7%

Context Poisoning

Guardrail Bypass

Privilege Escalation

Data Exfiltration

ctf-misc/games-and-vms.mdMEDIUM
86.1%

Tool passed security scan

Data Exfiltration

Context Poisoning

Guardrail Bypass

Risky tool definition detected

Tool: ctf-misc/games-and-vms.md [4/4] Description: # Merge: W' = W + B @ A merged = base['layer2.weight'] + lora['layer2.lora_B'] @ lora['layer2.lora_A'] # Threshold to binary image — values cluster at 0 or 1 binary = (merged > 0.5).int().numpy().astype(np.uint8) img = Image.fromarray((1 - binary) * 255) # Invert: 0→white, 1→black img.save('flag.png') ``` **Key insight:** LoRA adapters are low-rank matrix decompositions designed for fine-tuning. The product of the two small matrices can encode a

ctf-misc/dns.mdMEDIUM
79.9%

Malicious tool definition detected

ctf-misc/encodings.mdLOW
68.8%

Tool passed security scan

ctf-misc/bashjails.mdLOW
57.3%

Tool passed security scan

Passed Files (2)Click to expand
ctf-misc/rf-sdr.mdOK
29.0%

Tool passed security scan

ctf-misc/pyjails.mdOK
19.1%

Tool passed security scan

Audit Metadata
Max File Score
90%
Classification
KNOWN_SERVER_PARTIAL_KNOWN
Files Scanned
7
Files Flagged
5
Chunks Analyzed
13
Analyzed
Mar 16, 2026, 07:11 AM
Security Audit — runlayer — ctf-misc