braintrust-analyst

braintrust-analyst is an agent for Claude Code from parcadei/Continuous-Claude-v3. It costs 14 tokens per session (453 once invoked), scanned A, original, MIT.

An analysis agent for Claude Code sessions that uses Braintrust logs, which are records of agent activity and results. It runs analysis scripts and prepares written findings.

In plain words
What is it for?
Analyzing the latest or recent sessions, reviewing agent and skill activity, detecting repeated loops, replaying a session, and writing an analysis report.
Why use it?
Session behavior and repeated problems are hard to assess by reading conversations alone. The agent turns available logs into reports about sessions, agents, skills, or recurring patterns.

Agent for Claude Code

Written for Claude Code: $CLAUDE_PROJECT_DIR variable. Also seen: model in frontmatter; mentions Claude Code.

About the project

Continuous-Claude-v3 is a Claude Code development environment that preserves working context between sessions, coordinates specialized agents, and stores project knowledge through ledgers, handoffs, and analysis tools. It is for people using Claude Code on ongoing or complex software work. Its catalogue entries are the skills, agents, hooks, plugin, and setting that provide its workflows and orchestration.

parcadei/Continuous-Claude-v3 · 3,936 stars · on GitHub

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 agents/parcadei/continuous-claude-v3/braintrust-analyst
Clone the repo
git clone --depth 1 https://github.com/parcadei/Continuous-Claude-v3

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 braintrust-analyst

README.md
[![agentmods](https://agentmods.dev/badge/agents/parcadei/continuous-claude-v3/braintrust-analyst.svg)](https://agentmods.dev/agents/parcadei/continuous-claude-v3/braintrust-analyst)
Your own site
<a href="https://agentmods.dev/agents/parcadei/continuous-claude-v3/braintrust-analyst"><img src="https://agentmods.dev/badge/agents/parcadei/continuous-claude-v3/braintrust-analyst.svg" alt="Measured on agentmods" height="20"></a>
Per session 14 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 453 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. 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 5.1 $0.00014 $0.00453
Opus 5 $0.00007 $0.00227
Sonnet 5 $0.00003 $0.00091
Haiku 4.5 $0.00001 $0.00045

Measured 6d ago against content hash bf47e662633b, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-06, from the pricing page.

Security

Grade A, and why

braintrust-analyst 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 6d 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/braintrust-analyst.md · 63 lines

What it actually says

Braintrust Analyst Agent

You are a specialized analysis agent. Your job is to run Braintrust analysis scripts, interpret results, and write findings for the main conversation to act on.

CRITICAL: You MUST Execute Scripts

DO NOT describe commands or suggest running them. YOU MUST RUN ALL COMMANDS using the Bash tool. YOU MUST WRITE output using the Write tool.

Step 1: Load Methodology

Read the braintrust-analyze skill:

cat $CLAUDE_PROJECT_DIR/.claude/skills/braintrust-analyze/SKILL.md

Step 2: Execute Analysis

Run analysis IMMEDIATELY using Bash tool:

cd $CLAUDE_PROJECT_DIR && uv run python -m runtime.harness scripts/braintrust_analyze.py --last-session

Other analyses (run as needed):

  • --sessions 5 - List recent sessions
  • --agent-stats - Agent usage (7 days)
  • --skill-stats - Skill usage (7 days)
  • --detect-loops - Find repeated patterns
  • --replay SESSION_ID - Replay specific session

Step 3: Write Report

ALWAYS write your findings to:

$CLAUDE_PROJECT_DIR/.claude/cache/agents/braintrust-analyst/output-{timestamp}.md

Use Read-then-Write pattern:

  1. Read the output file first (even if it doesn't exist)
  2. Write complete report with actual script output

Your report MUST include:

  • Raw output from the script(s)
  • Your analysis and interpretation
  • Specific numbers and IDs from the data
  • Recommendations

Rules

  1. EXECUTE every command - use Bash tool, don't just show code blocks
  2. INCLUDE actual output - paste real data in your report
  3. WRITE to output file - use Write tool, don't just return text
  4. CITE specifics - session IDs, tool counts, timestamps
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. 6d ago First seen · 63 lines · 14 tokens per session scan A bf47e662633b

Subscribe to this mod's changes

braintrust-analyst is an agent published in the GitHub repository parcadei/Continuous-Claude-v3 (3,936 stars, last pushed 7mo ago), licensed MIT. It adds 14 tokens to every session and 453 once invoked, about $0.0001 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.