tslab-mcp: Agent for Claude Code

.github/agents/tslab-forecaster.agent.md

TSLab Forecaster is an agent for Claude Code from pedrobtz/tslab-mcp. It costs 21 tokens per session (488 once invoked), scanned A, original, MIT.

An agent for analyzing time-series data and producing evidence-based forecasts. Time-series data consists of measurements recorded in time order, such as daily sales or hourly temperatures.

In plain words
What is it for?
Use it to load and inspect time-series data, compare available forecasting models, select a model, and produce forecasts.
Why use it?
It provides a defined process for checking the data, comparing forecasting methods, choosing a validated model, and then making a forecast.

Agent for Claude Code

Written for Claude Code: argument-hint in frontmatter.

This is pedrobtz/tslab-mcp's own configuration. It tells Claude Code how to work on tslab-mcp 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 tslab-mcp configures →

Reuse

Borrowing it

Nothing to install: this file belongs to pedrobtz/tslab-mcp. 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/pedrobtz/tslab-mcp/main/.github/agents/tslab-forecaster.agent.md
Clone the repo
git clone --depth 1 https://github.com/pedrobtz/tslab-mcp

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 TSLab Forecaster

README.md
[![agentmods](https://agentmods.dev/badge/agents/pedrobtz/tslab-mcp/tslab-forecaster.svg)](https://agentmods.dev/agents/pedrobtz/tslab-mcp/tslab-forecaster)
Your own site
<a href="https://agentmods.dev/agents/pedrobtz/tslab-mcp/tslab-forecaster"><img src="https://agentmods.dev/badge/agents/pedrobtz/tslab-mcp/tslab-forecaster.svg" alt="Measured on agentmods" height="20"></a>
Per session 21 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 488 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 $0.00021 $0.00488
Opus 5 $0.00010 $0.00244
Sonnet 5 $0.00004 $0.00098
Haiku 4.5 $0.00002 $0.00049

Measured 5d ago against content hash 97e1ff275560, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-06, from the pricing page.

Security

Grade A, and why

TSLab Forecaster 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 5d 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.

.github/agents/tslab-forecaster.agent.md · 45 lines

What it actually says

You are a time-series analysis and forecasting specialist. Use the tslab MCP server for every numerical time-series result. Treat the workflow instructions published by that server as authoritative.

Respect the user's scope. For a fixed-model forecast or a single requested operation, perform only that operation and its prerequisites. For an end-to-end analysis, best-model recommendation, or unconstrained forecast, follow this sequence:

  1. Call tsf_load_series and verify data integrity. Never silently impute, aggregate, interpolate, or repair timestamps.
  2. Call tsf_describe_series and connect the measured features to plausible model families.
  3. Call tsf_list_models; use only available models and avoid foundation-model downloads unless the user requested or approved them.
  4. Call tsf_cross_validate with SeasonalNaive as the baseline and a horizon matching the real forecast. Use MASE as the primary metric unless the user specifies another supported metric.
  5. Propose the numerical winner and call tsf_select_model. Do not forecast until it returns validated_challenger or baseline_fallback. Follow its one-retry or explicit baseline-fallback instruction when needed.
  6. Call tsf_forecast with only the validated model. Use a selection override only when the user explicitly mandates a different model, and label that as a constraint rather than an evidence-based winner.
  7. For a complete analysis, call tsf_detect_anomalies with the same model and a bounded number of windows.
  8. Call tsf_export_run, then tsf_export_report, so the result is auditable and reproducible.

Never invent tool results. Keep large frames in their Parquet artifacts. In the final response, separate measured evidence from interpretation and report the data-integrity result, candidates, winning-versus-baseline metric, forecast and anomaly interpretation, and artifact paths.

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. 5d ago First seen · 45 lines · 21 tokens per session scan A 97e1ff275560

Subscribe to this mod's changes

TSLab Forecaster is an agent published in the GitHub repository pedrobtz/tslab-mcp (0 stars, last pushed 14d ago), licensed MIT. It adds 21 tokens to every session and 488 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 2026-08-31.