feat(skills): add SNMP enumeration and exploitation skill - #997
feat(skills): add SNMP enumeration and exploitation skill#997sandiyochristan wants to merge 2 commits into
Conversation
Greptile SummaryThe PR adds an SNMP protocol skill covering discovery, community-string testing, MIB enumeration, write-access validation, SNMPv3 weaknesses, and exploitation workflows. The current guidance still leaves autonomous target state modified and omits a prerequisite for one documented command.
Confidence Score: 2/5The PR should not merge until autonomous write validation restores target state and every unavailable documented tool has usable prerequisite guidance. The skill directs autonomous agents to leave sysContact overwritten after validation, and the outstanding prerequisites issue still causes the documented Hydra command to fail in the standard sandbox. Files Needing Attention: strix/skills/protocols/snmp.md
|
| Filename | Overview |
|---|---|
| strix/skills/protocols/snmp.md | Adds the SNMP skill, but autonomous write validation lacks rollback and the tooling section still omits Hydra's prerequisite. |
Prompt To Fix All With AI
### Issue 1
strix/skills/protocols/snmp.md:138-139
**Write validation leaves persistent state**
When an autonomous agent validates a writable SNMP endpoint, this sequence replaces the device's existing `sysContact` value with `strix_write_test` but never preserves or restores the original value, leaving administrative contact information modified after the scan. **How this was verified:** The autonomous validation sequence contains only the write and confirmation commands, and no applicable agent guidance supplies a restoration step.
### Issue 2
strix/skills/protocols/snmp.md:226
**Hydra prerequisite remains undocumented**
When an agent selects the documented Hydra community-string guessing command, the standard sandbox has no Hydra binary and this prerequisites section provides no installation instruction, so the command terminates with `hydra: command not found` instead of testing the target.
---
For each issue above, determine whether it is valid and should be fixed. If so, fix it directly.Reviews (2): Last reviewed commit: "fix(skills): address Greptile review fee..." | Re-trigger Greptile
![greptile-apps[bot]](https://avatars.githubusercontent.com/in/867647?s=60&v=4){kind=small}
Covers community string guessing, MIB walking, write access abuse, SNMPv3 weaknesses, credential extraction, and trap abuse. Closes usestrix#993
- Move operator approval warning BEFORE destructive SET examples so agents encounter the gate before the commands - Add safe autonomous write confirmation using benign sysContact - Fix authNoPriv description: scoped PDU data is unencrypted, but auth passwords are NOT transmitted in plaintext (HMAC-based) - Add Tooling section documenting prerequisites (snmp package, onesixtyone) and noting nmap NSE scripts need no installation
4632250 to
504615e
Compare
| snmpset -v2c -c <write_community> <target> 1.3.6.1.2.1.1.4.0 s "strix_write_test" | ||
| snmpget -v2c -c <read_community> <target> 1.3.6.1.2.1.1.4.0 # Confirm change |
There was a problem hiding this comment.
When an autonomous agent validates a writable SNMP endpoint, this sequence replaces the device's existing sysContact value with strix_write_test but never preserves or restores the original value, leaving administrative contact information modified after the scan. How this was verified: The autonomous validation sequence contains only the write and confirmation commands, and no applicable agent guidance supplies a restoration step.
Knowledge Base Used: Skills Library
Prompt To Fix With AI
This is a comment left during a code review.
Path: strix/skills/protocols/snmp.md
Line: 138-139
Comment:
**Write validation leaves persistent state**
When an autonomous agent validates a writable SNMP endpoint, this sequence replaces the device's existing `sysContact` value with `strix_write_test` but never preserves or restores the original value, leaving administrative contact information modified after the scan. **How this was verified:** The autonomous validation sequence contains only the write and confirmation commands, and no applicable agent guidance supplies a restoration step.
**Knowledge Base Used:** [Skills Library](https://app.greptile.com/strix-org-3/-/custom-context/knowledge-base/usestrix/strix/-/docs/skills-library.md)
---
For each issue above, determine whether it is valid and should be fixed. If so, fix it directly.|
|
||
| ## Tooling | ||
|
|
||
| The Strix sandbox includes `nmap` (with NSE scripts). Additional tools may need installation: |
There was a problem hiding this comment.
When an agent selects the documented Hydra community-string guessing command, the standard sandbox has no Hydra binary and this prerequisites section provides no installation instruction, so the command terminates with hydra: command not found instead of testing the target.
Knowledge Base Used: Skills Library
Prompt To Fix With AI
This is a comment left during a code review.
Path: strix/skills/protocols/snmp.md
Line: 226
Comment:
**Hydra prerequisite remains undocumented**
When an agent selects the documented Hydra community-string guessing command, the standard sandbox has no Hydra binary and this prerequisites section provides no installation instruction, so the command terminates with `hydra: command not found` instead of testing the target.
**Knowledge Base Used:** [Skills Library](https://app.greptile.com/strix-org-3/-/custom-context/knowledge-base/usestrix/strix/-/docs/skills-library.md)
---
For each issue above, determine whether it is valid and should be fixed. If so, fix it directly.Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time!
2 participants
No description provided.