10 actions in 8 pieces of software · ranked by words, meaning and measured trust · 540 ms
modelrecord
- zf-kbot/image-background-remove
modelrecord
- zf-kbot/comfyui-text-unblur
modelrecord
- zf-kbot/sd-inpaint
Fill in masked parts of images with Stable Diffusion
- zf-kbot/sd-inpaint
modelrecord
- zf-kbot/live-portrait
live portrait with audio
- zf-kbot/live-portrait
modelrecord
- zf-kbot/photo-to-anime
Convert images to anime style
- zf-kbot/photo-to-anime
modelrecord
- zf-kbot/inpaint-and-guess-prompt
Use a mask to inpaint the image or generate a prompt based on the mask.
- zf-kbot/inpaint-and-guess-prompt
modelrecord
- zf-kbot/invsr
- 8zeroforgefound in code, not yet checkedno passport yet
- zf_exploit
Attempt an autonomous end-to-end exploit (currently ret2win for a no-canary stack overflow): dynamically compute the RIP offset, locate a win/system function, build a stack-aligned payload, run the target and prove code execution. Returns the result and the shell proof.
- zf_ingest
Characterize a binary and populate the knowledge graph: file type, arch, mitigations (checksec), dangerous sinks, interesting strings, and — when NX is on — ROP gadgets. Run this first on any new target.
- zf_assess
Narrative summary of the engagement: target profile, mitigations, discovered sinks, implied exploit primitives, and the most promising path.
- zf_exploit
Labels are the instruments' verdict and are never changed by ranking. Agents and robots get this same answer as JSON at /api/v1/find?q=…&answer=1 and through the MCP server.