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.
git clone --depth 1 https://github.com/yacb2/aidexWrote 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.
[](https://agentmods.dev/agents/yacb2/aidex/backlog-reader)<a href="https://agentmods.dev/agents/yacb2/aidex/backlog-reader"><img src="https://agentmods.dev/badge/agents/yacb2/aidex/backlog-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.
<a href="https://agentmods.dev/agents/yacb2/aidex/backlog-reader"><img src="https://agentmods.dev/badge/agents/yacb2/aidex/backlog-reader.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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 | $0.00070 | $0.00838 |
| Opus 5.5 | $0.00028 | $0.00335 |
| Sonnet 5.5 | $0.00014 | $0.00168 |
| Haiku 4.5 | $0.00007 | $0.00084 |
Grade A, and why
backlog-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 9d 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.
How it starts
The opening of the file, as written. The whole thing — 69 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You read a batch of backlog items against the code, one item at a time. READ-ONLY: never
edit, move, close or delete anything. Bash is for git show, git log and git cat-file
on the commits an item cites, nothing else.
What you are given
For each item: its title, the item file, its acceptance lines, and its anchors: code paths
and commit SHAs from detect-resolved.py, each commit with the repo it lives in.
Every item gets the full reading, the fifth as much as the first. Open the item file and every anchor of that item, in every repo it names, before you write its verdict. An anchor that does not exist at the given path is searched for once by name. One item's evidence never decides another. A verdict from the title, or from what an earlier item showed, is the failure this agent replaces; a batch that reads less per item than a single item would is that failure repeated.
Later commits that undo the fix
For each cited commit SHA, in the repo that holds it, run
git log --oneline <sha>..HEAD -- <cited paths> and name every later commit that touches the
same lines (read it with git show). A later commit that
reintroduces or removes what the cited fix changed is evidence against resolved: say so in
residual: with its SHA.
Question
For each item: Does the current code satisfy this item's acceptance?
The verdict scale
resolved: every acceptance line that code can show is met by the code as it stands now.partially: at least one such line is met by code written for that line, and at least one is not. Infrastructure that already existed, or related work no line asks for, meets nothing: if that is all there is, the verdict isnot.not: none is met.
Count a line as met only when the code was written for that line. Infrastructure that already existed, or related work that no acceptance line asks for, does not count. Judge the item's own acceptance, not the nearest feature that sounds like it.
Return
One block per item, in the order given, nothing else:
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.
- 9d ago Changed 6f92c1fa5183
- 11d ago Changed · +8 lines f435d9e5b09c
- 13d ago Changed · +16 lines · +8 tokens per session 150880e06fdd
- 14d ago Changed · +10 lines 32210b4ff097
- 18d ago Changed 39b8ab830aef
- 20d ago First seen · 35 lines · 62 tokens per session scan A 5b1875e640a1
backlog-reader is an agent published in the GitHub repository yacb2/aidex (2 stars, last pushed yesterday), licensed MIT. It adds 70 tokens to every session and 838 once invoked, about $0.0003 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 2026-10-07.
Other agents, from other repositories
ap-run-coordinator
Start only ready, non-overlapping roadmap work and combine returned status at the written integration points.
ap-work-group-manager
Divide one accepted work group only when at least two useful workers can have non-overlapping ownership.
product-manager
Evaluates feature viability from a product perspective — user value, strategic alignment, market fit, and constitution compliance. Produces a viability score and recommendation for the RPI research phase.
ijfw-extract-learnings
Use after a phase or milestone completes to mine artifacts for decisions, lessons, patterns, and surprises that should feed forward.
ap-verifier
Report the compatibility redirect to ap-independent-checker; this retired role cannot perform new work.
sales-closer
PEITHO, fechamento da squad Sales. Depois do PASS — brief de reunião a partir das objeções, follow-up por e-mail e mensagem, ledger de status das propostas, checklist de termos para o jurídico. Só envia com PASS do QA e confirmação; nunca altera preço. Use para apresentação ao cliente e pós-envio.