agentmods.dev

searchflow

searchflow is a skill for Claude Code, Codex from treylom/deep-research. It costs 73 tokens per session (7,089 once invoked), scanned B, original, MIT.

A research process for answering questions with sourced and cross-checked findings.

In plain words
What is it for?
It helps with fact-checking, market, company, technology, and policy research, comparisons, and open-ended investigations.
Why use it?
It adjusts the amount of investigation to the difficulty of the question and checks weak parts of the evidence before reaching a conclusion.

Skill for Claude CodeCodex

Part of the searchflow plugin — 1 skill, 1 command, 1 MCP server shipped together

not rated 2 changed today B scan 73 tokens original MIT
View the source file treylom/deep-research
Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add skills/treylom/deep-research/searchflow
Any agent
npx skills add treylom/deep-research --skill searchflow
Clone the repo
git clone --depth 1 https://github.com/treylom/deep-research

Made for: Claude Code, Codex.

Or install searchflow, the plugin that ships this one along with the rest of its 1 skill, 1 command, 1 MCP server.

Wrote 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.

agentmods badge for searchflow

README.md
[![agentmods](https://agentmods.dev/badge/skills/treylom/deep-research/searchflow.svg)](https://agentmods.dev/skills/treylom/deep-research/searchflow)
Your own site
<a href="https://agentmods.dev/skills/treylom/deep-research/searchflow"><img src="https://agentmods.dev/badge/skills/treylom/deep-research/searchflow.svg" alt="Measured on agentmods" height="20"></a>
Per session 73 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 7,089 The whole file, excluding the scripts and references it only reads on demand.
Security scan B 1 finding. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5 $0.00073 $0.07089
Opus 5 $0.00036 $0.03544
Sonnet 5 $0.00015 $0.01418
Haiku 4.5 $0.00007 $0.00709

Measured today against content hash 178809e37af7, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade B, and why

searchflow scanned grade B with 1 finding 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 today.

The scan reads SKILL.md. This mod also ships 13 executable files (fixtures/ac9-degrade-bare.sh, fixtures/s2-spawn-smoke.sh, scripts/ac6-compare.mjs, …), listed below but not scanned — reading those needs a real analyzer, not pattern matching.

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.

Reads agent configuration directoriesmediumAgent snooping

.claude/, .codex/, .gemini/ hold keys, settings and other credentials a mod has no legitimate need for.

`--no-install`(=`SEARCHFLOW_PREFLIGHT_INSTALL=0`) 은 점검만 하고, `SEARCHFLOW_PREFLIGHT_ENHANCED=0` 은 강화 요소 설치를 생략하며, `SEARCHFLOW_NODE` 는 실행기를 수동 지정한다. codex 등록 대상은 `~/.codex/config.toml` 이고, `CODEX_HOME` 이 설정돼 있으면 그 아래 `confi
skills/searchflow/SKILL.md · 329 lines

How it starts

The opening of the file, as written. The whole thing — 329 lines — stays where its author put it; the contents beside it link to each section on GitHub.

SearchFlow

소스 등급 가중 점수 + 은닉 문턱 게이트로 깊이를 사후 결정하는 리서치 공정.

깊이를 사용자에게 미리 묻지 않는다. 쉬운 질의는 1라운드에서 문턱을 넘어 즉시 끝나고, 어려운 질의만 재조사를 돈다 — 평균은 내려가고 하한은 올라간다.

0. 두 층

  • core (기본): 순정 Claude Code 또는 순정 Codex CLI + 이 스킬 번들만으로 완주한다. 외부 패키지 · 내부 MCP · 특정 플러그인 requirement 0. node 실행기는 scripts/preflight.sh(§5.1) 가 확보한다. 없으면 관리자 권한 없이 내려받고, 그래도 못 받으면 degraded=no-node-runtime 라벨을 남기고 완주한다.
  • enhanced (선택): 환경에 있으면 얹는다. 없으면 에러 없이 core 로 강등하고 격하 라벨을 보고서에 남긴다 — 조용한 skip 금지.

scripts/preflight.sh(§5.1) 다음에 scripts/env-detect.mjs 를 1회 실행해 참고값을 얻는다(인자 없이 호출 = exit 0 고정, stdout 1줄 JSON) — 호출은 preflight JSON 의 node 경로로. 탐지 결과는 리드만 소비한다.

⚠️ multi_agent_api 값은 참고다. multi-agent 경로의 최종 판정은 리드가 자기 tool 목록을 보고 한다 — shell 스크립트가 모델의 tool 노출을 대신 판정하면 틀린다.

0-1. 공정 소유권 — 스킬이 쥐나, 서버가 쥐나

위 두 층과 직교하는 축이다(core/enhanced = 어떤 도구를 얹나 · 아래 = 누가 공정을 강제하나).

모드 조건 채점 기준이 사는 곳 판정
skill-owned (기본) searchflow_* 도구가 tool surface 에 없다 references/scoring.md — 리드가 읽고 리드가 채점 리드
server-owned searchflow_start·searchflow_submit·searchflow_gate 가 tool surface 에 있다 MCP 서버 내부(파일로 존재하지 않음) 서버
  • 판정 주체 = 리드가 자기 tool 목록을 보고 한다. 위 multi_agent_api 주의와 같은 이유로 shell 이 대신 판정하지 않는다. 서버가 안 떠 있으면 도구가 애초에 안 보이므로 탐지 실패 = skill-owned 로 fail-open(에러 ❌).
  • server-owned 면 리드는 scoring.md 를 열지 않는다. 프레임·브리핑·판정을 전부 서버에서 받고, 서버가 준 worker_brief 를 그대로 워커에게 넘긴다(문장을 늘리지 않는다 — §P2.5 와 같은 이유).
  • 어느 쪽이든 보고서에 모드를 적는다: process_owner: skill | server. 조용한 전환 금지 — 같은 질의가 다른 깊이로 끝났을 때 원인이 모드였는지 알 수 있어야 한다.

⚠️ 이 표는 2026-08-02 신설이다. 그전까지 서버는 존재했지만 스킬이 그 존재를 몰랐다 — 즉 「2모드」가 설계 문서에만 있고 실행 경로엔 없었다. 두 모드를 비교하려면 먼저 모드가 있어야 한다.

1. 공정 (P0 → P4)

P0 스코핑 ── 유형 판정 → 프레임 확정 → 깊이 라우팅 → 묶음 질문 1회(마감 시한 포함) → 권한 preflight
P1 스폰   ── 프레임 축마다 조사 단위 1개 (하네스별 분기 §3)
P2 수집   ── 워커: 근거 + SELF-REPORT(Y/N + 근거 1줄)만 반환. 등급 부여 ❌
P3 채점   ── 리드: 소스 등급 부여 → 원장 기록 → 축별 채점 → 문턱 대조   ← 리드 전용
P4 게이트 ── 문턱 통과 → 종료 / 미달 → 최저 축만 재조사(P1 로) → 최대 2라운드

Read the full file on GitHub · 329 lines

Files

What ships with it

31 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.

Changes

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.

  1. today Changed · +27 lines scan A → B 178809e37af7
  2. 4d ago First seen · 302 lines · 73 tokens per session scan A 42eca2d1af97

Subscribe to this mod's changes

searchflow is a skill published in the GitHub repository treylom/deep-research (2 stars, last pushed yesterday), licensed MIT. It adds 73 tokens to every session and 7,089 once invoked, about $0.0004 per session on Opus 5. A static security scan graded it B with 1 finding (reads agent configuration directories). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.

Researchclaude-codecodexfact-checkingskills
Related

Other skills, from other repositories

belief-assessor

LLM-driven epistemic reasoning engine. Evaluates claims against evidence, outputs calibrated confidence and structured belief state (VERIFIED/CONTESTED/UNCERTAIN). v2 adds 4-way constraint system, parameterized configuration, and formula-based confidence intervals. Use when the agent needs to assess whether…

hqzzdsda/belief-state-runtime · 77 tokens

belief-state-runtime

LLM-driven epistemic reasoning engine. Evaluates claims against evidence, outputs calibrated confidence and structured belief state (VERIFIED/CONTESTED/UNCERTAIN). Use when the agent needs to assess whether information is trustworthy, detect contradictions in evidence, or quantify uncertainty.

hqzzdsda/belief-state-runtime · 58 tokens

claude-md-improver

Audit and improve CLAUDE.md files in repositories. Use when user asks to check, audit, update, improve, or fix CLAUDE.md files. Scans for all CLAUDE.md files, evaluates quality against templates, outputs quality report, then makes targeted updates. Also use when the user mentions "CLAUDE.md maintenance" or "project…

anthropics/claude-plugins-official · 82 tokens

agent-platform-rag-engine-management

Manage and query Agent Platform RAG Engine Corpora and retrieve grounded contexts using the Google GenAI SDK. Use when listing RAG corpora or files, inspecting a corpus, retrieving contexts, or generating content grounded in a RAG corpus. Do not use for standard database queries (use SQL/Spanner skills), Google…

google/skills · 85 tokens

agent-platform-model-registry

Agent Platform Model Registry Management. Use when you need to upload, list, describe, update, or delete machine learning models (and their versions) in the Agent Platform Model Registry. Don't use for model training, model deployment to endpoints, or managing non-Agent Platform models.

google/skills · 60 tokens

twitter-reader

Read Twitter/X for financial research using opencli (read-only). Use this skill whenever the user wants to read their Twitter feed, search for financial tweets, view bookmarks, look up user profiles, or gather market sentiment from Twitter/X. Triggers include: "check my feed", "search Twitter for", "show my…

himself65/finance-skills · 161 tokens