Getting it into your agent
It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.
git clone --depth 1 https://github.com/Encod3d-Sec/TORCHnpx agentmods add skills/encod3d-sec/torch/screenshot-burpWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/encod3d-sec/torch/screenshot-burp)<a href="https://agentmods.dev/skills/encod3d-sec/torch/screenshot-burp"><img src="https://agentmods.dev/badge/skills/encod3d-sec/torch/screenshot-burp/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/encod3d-sec/torch/screenshot-burp"><img src="https://agentmods.dev/badge/skills/encod3d-sec/torch/screenshot-burp.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 3 findings, up to medium
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- medium Rogue Agent · line 16 Skill establishes unauthorized persistence across sessions via cron jobs, startup scripts, or state files. Session persistence allows an attacker to maintain access beyond the current interaction.Fix: Remove any persistence mechanisms (cron jobs, startup scripts, state files). Skills should not maintain state across sessions without explicit user consent.
- medium Tool Misuse · line 42 Tool defaults are unsafe or overly permissive (e.g. disabled TLS verification, no authentication, world-writable permissions). Unsafe defaults widen the attack surface.Fix: Override unsafe defaults with secure settings (verify=True, auth required, restrictive permissions). Review and harden all tool configurations.
- medium Privilege Escalation · line 83 Commands invoke sudo or root privileges. Verify this elevated access is necessary and justified.Fix: Avoid sudo/root unless strictly required. Prefer least-privilege patterns. If elevation is needed, document the justification and scope.
What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00115 | $0.02245 |
| Opus 5 | $0.00057 | $0.01123 |
| Sonnet 5 | $0.00023 | $0.00449 |
| Haiku 4.5 | $0.00012 | $0.00225 |
Grade B, and why
screenshot-burp scanned grade B with 2 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 9d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Asks for rootmediumPrivilege escalation
A mod that escalates privileges can change anything on the machine, not only the project.
sudo -u "$U" env DISPLAY="$D" XAUTHORITY=/home/$U/.Xauthority bash -c " Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
description: Capture a Burp Suite Repeater request/response as a PoC image (targets/<eng>/poc/) by driving the Burp MCP + the Kali GUI. Replays a request in a Repeater tab, sends it, and grabs the request+response panes How it starts
The opening of the file, as written. The whole thing — 98 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Screenshot: Burp Repeater request/response PoC
Turn a Burp Repeater exchange into a report-ready request + response image. This is the Burp-native
counterpart to Skill(screenshot) (curl/terminal cards): when you drive a target through Burp, capture
the proof FROM Burp. Prereqs are the same as Skill(hunt-burp) (Burp running + the MCP Server BApp, SSE
on 127.0.0.1:9876; verify with bash /root/vm.sh 'python3 ~/burp-mcp-cli.py list').
One command (use this)
scripts/capture.sh burp <eng> <slug> <host> <port> <https:true|false> <method> <path> [bodyfile] [tabname]
# POST with a body file (forged/complex bodies -> write to a file, avoids shell quoting):
scripts/capture.sh burp thm_x flag1 10.1.1.1 5000 true POST /login /tmp/login_body.txt "ECorp login"
# simple GET:
scripts/capture.sh burp thm_x flag3 10.1.1.1 3000 false GET /challenge/solve
It: (1) create_repeater_tab via the MCP with the raw request, (2) activates Burp on the seat display,
focuses the request editor, sends with Ctrl+Space (Burp's Repeater Send hotkey), (3) import-grabs
the window and pulls targets/<eng>/poc/NN-<slug>.png, printing the ![]() ref. Drop that ref into
walkthrough.md so the image actually renders in Obsidian (an un-referenced PNG in poc/ is invisible
in the notes - only reachable by opening the folder).
Crypto-forged / signed requests: give the exploit script a --curl-style mode that writes the exact
body to a file (e.g. /tmp/login_body.txt), then pass it as bodyfile. The Repeater tab then holds the
real forged request, so the PoC is Burp-native and reproducible.
Gotchas baked into capture.sh burp (the burp mode) (why hand-rolling this failed the first time)
- Grab as the SEAT user, not root. Burp may be root-owned but it DRAWS on the desktop user's X
session (the desktop login on
:0).who-> the line with a(:N)display gives the user + display; use their~/.Xauthority. A root-over-SSH grab has no X display. - Send = Ctrl+Space (KEYBOARD only); mouse is dead. Java Swing IGNORES synthetic
xdotoolMOUSE clicks (button/tab/pixel clicks are silent no-ops), but KEYBOARD events land once the window is activated (windowactivate --sync):Ctrl+Shift+R(focus Repeater),Ctrl+=(next sub-tab),Ctrl+Space(Send). Drive everything by keyboard; anything with no hotkey (e.g. the BAppMCPtab) needs a human mouse action. - Window offset:
getwindowgeometry-> the client area sits at screen (0, ~35); theimport -windowgrab starts at the client top, so screen_y = image_y + 35 (only matters if you ever DO need a click on a WM that accepts synthetic clicks; this one does not). - Write the grab to a WORLD-WRITABLE path (
/tmp/capture_burp_*.png), never a root-owned-odir: the sudo-as-seat-userimportcan't write into/tmp/pocif root created it (silent EACCES = "no PNG"). - The "SSE wedge" was a MISDIAGNOSIS (corrected 2026-07-24). The only MCP call that ever hung is
send_http1_request, and that is the extension's target-approval gate (a non-approved target raises a GUI approval prompt a headless seat cannot answer -> 15s timeout), NOT a per-session wedge.create_repeater_tab,get_active_editor_contents,url_encode,set_proxy_intercept_stateall run fine across many back-to-back per-call sessions.capture.sh burpnever callssend_http1_request(it Sends in the GUI viaCtrl+Space, human-equivalent, which bypasses the approval gate), so it is unaffected. Ifcreate_repeater_tabitself fails, the server is down/unreachable -> checkscripts/burp/burp-transport.sh(see [[burp-mcp]]).
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 9d ago First seen · 98 lines · 115 tokens per session scan B d4d1bb5d65f8
screenshot-burp is a skill published in the GitHub repository Encod3d-Sec/TORCH (318 stars, last pushed 8d ago), licensed MIT. It adds 115 tokens to every session and 2,245 once invoked, about $0.0006 per session on Opus 5. A static security scan graded it B with 2 findings (asks for root, makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
systematic-debugging
Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.
local-ai-agents
Build local-first AI agents that run entirely on a developer workstation with Microsoft Foundry Local and Qwen function-calling models. Covers Small Language Models (SLMs), the OpenAI-compatible local endpoint, sandboxed local tools, local RAG with Chroma, local MCP servers, hybrid cloud/local routing, and the…
next-cache-components-adoption
Turn on Cache Components in a Next.js app and resolve the blocking routes it surfaces. Use when the user wants to enable, adopt, or migrate to Cache Components, flip the cacheComponents flag, work through a flood of blocking-prerender / instant validation errors, run the cache-components-instant-false codemod, or…
insight-error-page
Write or audit an insight-kind error page for the Next.js dev overlay. Use when creating a new errors/ .mdx page, auditing an existing one, or checking that a page matches the framework fix cards. Covers page structure, title alignment, FixCard cards with Copy prompt button, code snippets, terminology verification…
next-cache-components-optimizer
Drive a Next.js route to instant navigation by setting up an agentic loop, under Cache Components / PPR, on initial load (hard navigation) and client-side navigation (soft navigation). Encode the goal as a failing @next/playwright instant() e2e and work it to green, one verified route at a time; the shipped test then…
next-partial-prefetching-adoption
Turn on Partial Prefetching in a Next.js app and work through the insights it surfaces. Use when the user wants to enable or adopt Partial Prefetching, flip the partialPrefetching flag, opt routes in with export const prefetch = 'partial', audit Link prefetch={true} behavior, preserve existing prefetched UI with…