What is session-reader?

A session-analysis agent that reads Claude Code session logs, which are JSONL files containing one JSON record per line, and summarizes the work done.

Use it to analyze one or more coding sessions, especially when several session files need to be summarized in parallel.

It removes tool-call noise and turns long session files into a readable record of requests, decisions, technologies, and outcomes.

How GengzeZhou/memex configures Claude Code

Reuse

Borrowing it

Nothing to install: this file belongs to GengzeZhou/memex. 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/GengzeZhou/memex/main/.claude/agents/session-reader.md

Clone the repo

git clone --depth 1 https://github.com/GengzeZhou/memex

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.

README.md

[![agentmods](https://agentmods.dev/badge/agents/gengzezhou/memex/session-reader/github.svg)](https://agentmods.dev/agents/gengzezhou/memex/session-reader)

Your own site

<a href="https://agentmods.dev/agents/gengzezhou/memex/session-reader"><img src="https://agentmods.dev/badge/agents/gengzezhou/memex/session-reader/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.

Your own site · 80×15

<a href="https://agentmods.dev/agents/gengzezhou/memex/session-reader"><img src="https://agentmods.dev/badge/agents/gengzezhou/memex/session-reader.svg" alt="Reviewed on agentmods" width="80" height="20"></a>

Per session 39 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.

When invoked 310 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.

Model	Per session	Once invoked
Fable 5.1	$0.00039	$0.00310
Opus 5.5	$0.00016	$0.00124
Sonnet 5.5	$0.00008	$0.00062
Haiku 4.5	$0.00004	$0.00031

Measured 2mo ago against content hash 3a6735af8958, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

session-reader 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 2mo 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/session-reader.md · 38 lines

What it actually says

You are a session analyst. Your job is to read a Claude Code session JSONL file and extract a structured summary.

Each line in the JSONL is a JSON object with fields:

type: "user", "assistant", "file-history-snapshot", etc.
message: the content (may be a string or a serialized dict)
cwd: working directory
sessionId: session identifier
timestamp: ISO timestamp or Unix ms

Focus on:

User messages (type: "user") — what was requested
Assistant messages (type: "assistant") — what was done
Skip: file-history-snapshots, tool call details, subagent notifications

Strip noise from user messages:

<ide_opened_file> tags — just note the filename
<local-command-stdout> — skip unless it contains key output
<command-name> tags — skip
<task-notification> — note only the summary line

Return a structured analysis with:

Project name and path
Date range
List of tasks attempted
Key decisions and outcomes
Technologies used
Any notable context

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.

2mo ago First seen · 38 lines · 39 tokens per session scan A 3a6735af8958

Subscribe to this mod's changes

session-reader is an agent published in the GitHub repository GengzeZhou/memex (10 stars, last pushed 7mo ago), licensed MIT. It adds 39 tokens to every session and 310 once invoked, about $0.0002 per session on Opus 5.5. 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 2026-08-31.

Building agents

session-reader next to 6 agents from other repositories
Mod	Type	Stars	Last push	Licence	Scan	Tokens	Works with
session-readerGengzeZhou/memex	Agent	10	7mo ago	MIT	A	39	Claude Code
playwright-test-generatormicrosoft/playwright	Agent	99k	yesterday	Apache-2.0	A	151	Claude Code
.NET-Notebook-Migration-Agentmicrosoft/ai-agents-for-beginners	Agent	79k	4d ago	MIT	A	33	Claude Code
code-revieweranthropics/claude-cookbooks	Agent	54k	2d ago	MIT	A	52	Claude Code
AVM Owner Triagegithub/awesome-copilot	Agent	41k	yesterday	MIT	A	61	Claude Code
Ultimate Transparent Thinking Beast Modegithub/awesome-copilot	Agent	41k	yesterday	MIT	A	11	Claude Code
WinForms Expertgithub/awesome-copilot	Agent	41k	yesterday	MIT	A	19	Claude Code

memex: Agent for Claude Code