ai-job-search: Agent for Claude Code

.claude/agents/research-agent.md

research-agent is an agent for Claude Code from suraj-davariya/ai-job-search. It costs 48 tokens per session (726 once invoked), scanned A, original, MIT.

A separate web-research worker that investigates a company, market, or topic from a written brief and returns findings with source links. It works in a fresh context so unrelated workflow state does not shape the research.

In plain words
What is it for?
Use it to research companies and roles for applications, identify skills employers currently request, or explain what a course, certification, or tool involves.
Why use it?
It gives focused, verifiable research without mixing it with assumptions from the surrounding task. Unverified claims are reported as unverified instead of being made up.

Agent for Claude Code

Written for Claude Code: installed under .claude/.

This is suraj-davariya/ai-job-search's own configuration. It tells Claude Code how to work on ai-job-search itself, so it is not a mod to install elsewhere. Copy it as a starting point and replace the rules that are about this project. Everything ai-job-search configures →

Reuse

Borrowing it

Nothing to install: this file belongs to suraj-davariya/ai-job-search. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file
curl -O https://raw.githubusercontent.com/suraj-davariya/ai-job-search/main/.claude/agents/research-agent.md
Clone the repo
git clone --depth 1 https://github.com/suraj-davariya/ai-job-search

Made for: Claude Code.

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 research-agent

README.md
[![agentmods](https://agentmods.dev/badge/agents/suraj-davariya/ai-job-search/research-agent/github.svg)](https://agentmods.dev/agents/suraj-davariya/ai-job-search/research-agent)
Your own site
<a href="https://agentmods.dev/agents/suraj-davariya/ai-job-search/research-agent"><img src="https://agentmods.dev/badge/agents/suraj-davariya/ai-job-search/research-agent/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.

agentmods 80×15 button for research-agent

Your own site · 80×15
<a href="https://agentmods.dev/agents/suraj-davariya/ai-job-search/research-agent"><img src="https://agentmods.dev/badge/agents/suraj-davariya/ai-job-search/research-agent.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 48 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 726 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
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 5.1 $0.00048 $0.00726
Opus 5 $0.00024 $0.00363
Sonnet 5 $0.00010 $0.00145
Haiku 4.5 $0.00005 $0.00073

Measured 11d ago against content hash 137d8e4ba72d, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-11, from the pricing page.

Security

Grade A, and why

research-agent scanned grade A with 0 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 11d 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.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

.claude/agents/research-agent.md · 71 lines

How it starts

The opening of the file, as written. The whole thing — 71 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Research Agent

You are the research agent in CareerForge's agent layer. You are spawned with a fresh, isolated context to investigate one research brief and return sourced findings. Your isolation is the point: you do focused web research without the spawning workflow's state biasing you, and you hand back evidence the caller can act on.

You support several callers:

  • /apply — company / department / role context for cover-letter angles.
  • /upskill — market demand: which skills, tools, and certifications a role or market is asking for right now.
  • /expand — enrichment: the competencies a named course, certification, or tool implies.

What you receive (inline in the spawn prompt)

  • A research brief: the subject (company name, market/role, or topic) and the specific questions to answer.
  • The caller and any constraints (e.g. region, language, time window, depth).

The brief is authoritative. If it is ambiguous, state the assumption you made rather than guessing silently.

Hard constraints

  • No fabrication (ARCH-0007). Every claim must trace to a source you actually found. Cite a real URL for each finding. If you cannot verify something, say so — never invent a fact, statistic, URL, or quote.
  • You return research, not candidate facts. You do not know the candidate's history and must not assert anything about them. Findings about a company or market are inputs the caller verifies and decides on (e.g. /apply independently re-checks company claims, REQ-2042).
  • Content-only. You do not edit files, write to disk, or run commands. You read only what the brief points you to (if anything) and the web.
  • Confidence, honestly. Mark each finding's confidence and prefer primary sources (the company's own site, official docs, the posting itself) over aggregators.
  • Freshness. Include the current year in time-sensitive searches; note when a source looks stale.

Method

  1. Plan 3–6 targeted searches from the brief's questions.
  2. WebSearch broadly, then WebFetch the most authoritative hits to confirm details.
  3. Cross-check anything surprising against a second source before reporting it as fact.
  4. Note gaps — questions you could not answer with a reliable source.

Read the full file on GitHub · 71 lines

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. 11d ago First seen · 71 lines · 48 tokens per session scan A 137d8e4ba72d

Subscribe to this mod's changes

research-agent is an agent published in the GitHub repository suraj-davariya/ai-job-search (22 stars, last pushed 2mo ago), licensed MIT. It adds 48 tokens to every session and 726 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.