원클릭으로 Manus에서 모든 스킬 실행

snitch

스타6

포크1

업데이트2026년 6월 26일 14:01

Inspect the exact bytes an LLM agent or SDK sends upstream — the full system prompt, every tool schema, and all messages — via a local transparent proxy with a live web UI. Use when debugging why an agent misbehaves, auditing hidden or injected prompt/tools, checking how much context a framework actually sends, or verifying the real request behind an SDK call.

설치

Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.

Manus에서 실행

출처

JGalego

JGalego/Bag-of-Tricks

GitHub 저장소 열기 Creator 저장소 보기

다운로드

Manus에서 실행

snitch

see what your agent actually said behind your back.

snitch is a transparent proxy that logs the exact request your code sends to an LLM API, then forwards it unchanged. Use it to reveal the system prompt, tool schemas, and injected context a framework hides.

When to use

"Why is my agent doing X?" → see the actual prompt/tools it was sent.
"Is something injecting instructions/tools?" → audit the raw request.
"How many tokens of system prompt / tools is this framework adding?"
"What does the SDK actually send for this call?"

How to run it

If installed on the PATH (via just install snitch):

snitch --port 8787            # starts proxy + web UI on http://localhost:8788

Otherwise run it from the bag-of-tricks repo: python3 snitch/snitch.py.

Then point the client at the proxy and run the agent normally:

export ANTHROPIC_BASE_URL=http://localhost:8787   # or OPENAI_BASE_URL, etc.

Console prints each request live; the web UI (http://localhost:8788) lets you scroll back, expand, and filter captured requests.
--upstream URL forwards to another provider; --log FILE appends each request as JSON; --quiet silences the console (UI + log still capture); --no-ui disables the web UI.

It captures the request in full and records the response status (it doesn't buffer response bodies). It binds to 127.0.0.1 only — a local debug tool.

Output

If you run snitch or report what it captured, deliver the captured request plainly — don't narrate the proxy mechanics or wrap the bytes in analysis the user didn't ask for, and point them at the web UI for the rest. No preamble, no closing summary.

See README.md in this folder for the full reference.

이 저장소의 다른 Skills

같은 저장소

alibi

JGalego/Bag-of-Tricks

Verify that an answer's claims are actually grounded in the provided sources, flagging fabricated or unsupported statements before you trust the output. Split the answer into sentence-level claims and check each one for lexical support in the source documents. Use after RAG or summarization, before trusting or shipping the result, to catch the sentence the model invented that the sources never said.

2026-06-266

bluff

JGalego/Bag-of-Tricks

Verify that the URLs and citations in an answer actually resolve, catching hallucinated or dead links before you trust or ship them. Extract every link (bare and markdown), check each over the network, and flag the ones that don't exist. Use when an LLM produced links/citations and you need to confirm they're real, not plausible-looking fiction.

2026-06-266

combo

JGalego/Bag-of-Tricks

Chain several tricks into one pipeline so the output of one flows straight into the next — redact then wash then strip, or wash then score — and run the whole routine in a single call. Use when a task needs more than one trick in sequence (clean-and-check, extract-and-analyze, gate-then-transform) instead of invoking each trick separately. Built on the Unix pipe: every trick is a stdin→stdout program, and combo wires the stages together, forwards each stage's summary, and aborts the moment a gate fails.

2026-06-266

fold

JGalego/Bag-of-Tricks

Catch overconfident phrasing in an answer and prefer calibrated uncertainty or honest abstention. Flag absolutes (always/never/guaranteed), bare certainty (definitely/obviously), and false authority (trust me / everyone knows) where the evidence is thin. Use when reviewing a draft answer for unwarranted certainty, or as a reflex to fold a weak hand — say "I don't know" — instead of confabulating a confident wrong answer.

2026-06-266

frisk

JGalego/Bag-of-Tricks

Pat text down for secrets and PII before it leaves the building. Scan prompts, context, and pasted snippets for API keys, tokens, private keys, JWTs, and emails — then redact or refuse. Use before sending context/prompts that may contain credentials to an LLM or any third party, before pasting logs into an issue, or before piping output anywhere it'll be stored.

2026-06-266

grill

JGalego/Bag-of-Tricks

Adversarially interrogate an answer before trusting it — generate probing follow-up questions that attack its weak points across hidden assumptions, missing edge cases, internal contradictions, unsupported claims ("what's the source?"), overconfidence, and "what would change your mind?", then optionally run them against a model to see whether the answer holds up or cracks. Use to stress-test a draft answer or another model's output before you ship or trust it.

2026-06-266

name	snitch
description	Inspect the exact bytes an LLM agent or SDK sends upstream — the full system prompt, every tool schema, and all messages — via a local transparent proxy with a live web UI. Use when debugging why an agent misbehaves, auditing hidden or injected prompt/tools, checking how much context a framework actually sends, or verifying the real request behind an SDK call.

snitch

see what your agent actually said behind your back.

When to use

"Why is my agent doing X?" → see the actual prompt/tools it was sent.
"Is something injecting instructions/tools?" → audit the raw request.
"How many tokens of system prompt / tools is this framework adding?"
"What does the SDK actually send for this call?"

How to run it

If installed on the PATH (via just install snitch):

snitch --port 8787            # starts proxy + web UI on http://localhost:8788

Otherwise run it from the bag-of-tricks repo: python3 snitch/snitch.py.

Then point the client at the proxy and run the agent normally:

export ANTHROPIC_BASE_URL=http://localhost:8787   # or OPENAI_BASE_URL, etc.

Console prints each request live; the web UI (http://localhost:8788) lets you scroll back, expand, and filter captured requests.
--upstream URL forwards to another provider; --log FILE appends each request as JSON; --quiet silences the console (UI + log still capture); --no-ui disables the web UI.

It captures the request in full and records the response status (it doesn't buffer response bodies). It binds to 127.0.0.1 only — a local debug tool.

Output

See README.md in this folder for the full reference.