01-metrics

01-metrics is an agent for coding agents from proggarapsody/bitbottle. It costs 0 tokens per session (3,769 once invoked), scanned A, original, MIT.

A metrics analysis of automated iteration cycles, checking whether timing, token, step, and blocker data is complete and trustworthy.

In plain words
What is it for?
Reviewing instrumentation quality, finding incorrect or absent fields, and understanding how long different cycle steps take.
Why use it?
It shows where measurement data is missing or inconsistent, so reported cycle costs and performance are easier to judge correctly.

Agent

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/proggarapsody/bitbottle/01-metrics
Clone the repo
git clone --depth 1 https://github.com/proggarapsody/bitbottle

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 01-metrics

README.md
[![agentmods](https://agentmods.dev/badge/agents/proggarapsody/bitbottle/01-metrics.svg)](https://agentmods.dev/agents/proggarapsody/bitbottle/01-metrics)
Your own site
<a href="https://agentmods.dev/agents/proggarapsody/bitbottle/01-metrics"><img src="https://agentmods.dev/badge/agents/proggarapsody/bitbottle/01-metrics.svg" alt="Measured on agentmods" height="20"></a>
Per session 0 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 3,769 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 $0.00000 $0.03769
Opus 5 $0.00000 $0.01885
Sonnet 5 $0.00000 $0.00754
Haiku 4.5 $0.00000 $0.00377

Measured 4d ago against content hash 7544bb425125, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

01-metrics 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 4d 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.

auto-iter/reports/analysis-2026-06-02-cycles-158-187/agents/01-metrics.md · 244 lines

How it starts

The opening of the file, as written. The whole thing — 244 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Metrics Agent Report — cycles 153–187 (2026-06-02)

Analysis dimension: METRICS — quality, completeness, and trustworthiness of auto-iter instrumentation; what the metrics reveal about cycle performance.


TL;DR

  • Data completeness is 37% overall (11/30 cycles have all five key fields populated); the dominant gap is the 168–177 zero-emission regime where tokens, duration, and metric steps were never written.
  • TDD dominates wall time at 84–92% of instrumented step time (cycles 178–187 median 9.6 min/cycle); design-judge consumes a consistent 8–17%; CI wait and PRD are not instrumented in the latest stream at all.
  • Design judge ran in 12/30 cycles and flagged blockers in 7 of the 11 cycles where blocker_count was recorded (64% blocker rate); all 7 were confirmed real (0% false-positive rate per stream-178-187 report).
  • Five distinct instrumentation defects exist: corrupt duration_wall_min on cycle 157 (29.6M min instead of 31 min); field-name inconsistency (blockers vs blocker_count); step-name drift across 4 canonical groups; metrics_steps_count mismatches on 4 cycles; and five cycles (162, 163, 165, 166, 167) entirely absent from both dataset and raw cycles.jsonl.

1. Data Completeness

Five key fields assessed: tokens > 0, duration_active_min > 0, release populated, scopes populated, _metric_steps non-empty. Brainstorm cycles excused from release requirement.

Regime Cycles tokens>0 duration>0 scopes steps>0 DJ step Fully complete
153–164 (pre-stream) 10 6/10 9/10 7/10 8/10 4 1/10 (10%)
168–177 (zero-logged) 10 0/10 0/10 10/10 1/10 0 0/10 (0%)
178–187 (instrumented) 10 9/10 9/10 10/10 10/10 8 10/10 (100%)
Total 30 15/30 18/30 27/30 19/30 12 11/30 (37%)

Five cycles completely absent (162, 163, 165, 166, 167) from both the dataset and cycles.jsonl — they have entries in metrics.jsonl only. This leaves a 5-cycle gap in any per-cycle lineage analysis. The metrics.jsonl entries for 162 and 163 show only step3_tdd + step8_ship (no CI wait, no DJ). Cycle 165 has a corrupt step2_prd duration (see §5).

Read the full file on GitHub · 244 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. 4d ago First seen · 244 lines · 0 tokens per session scan A 7544bb425125

Subscribe to this mod's changes

01-metrics is an agent published in the GitHub repository proggarapsody/bitbottle (3 stars, last pushed 17d ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 3,769 tokens. 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.