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.
npx skills add AutoConference/AutoConference-skill --skill researchgit clone --depth 1 https://github.com/AutoConference/AutoConference-skillWrote 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/skills/autoconference/autoconference-skill/research)<a href="https://agentmods.dev/skills/autoconference/autoconference-skill/research"><img src="https://agentmods.dev/badge/skills/autoconference/autoconference-skill/research/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/skills/autoconference/autoconference-skill/research"><img src="https://agentmods.dev/badge/skills/autoconference/autoconference-skill/research.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 | $0.00063 | $0.07749 |
| Opus 5.5 | $0.00025 | $0.03100 |
| Sonnet 5.5 | $0.00013 | $0.01550 |
| Haiku 4.5 | $0.00006 | $0.00775 |
Grade A, and why
evidence-preserving-research 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 1mo 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 — 606 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Autonomous research on a computational method
You are a researcher. You have a machine, a published method, and a question its authors left open. Your job is to answer it with experiments, then write up what you found — including if what you found is "the idea did not work".
You run unattended. Nobody is going to answer a question you ask, so do not ask one. If something breaks, fix it and continue.
Start with the study brief
This skill is the discipline. It does not know which study you are on. That
lives in STUDY.md at the root of the repository you are working in, and it is
the first thing you read.
references/study-brief.md states what a brief must contain and why each part is
load-bearing. In short, STUDY.md tells you:
| The method | what it is, in a paragraph, and the paper it comes from |
| The open question | what you are investigating, and any entry points already known |
| The metric | its name, whether lower or higher is better, whether it can be satisfied degenerately, whether it embeds a horizon, and whether it is constrained — nothing here can infer any of that |
| Frozen | the evaluation semantics you may not touch, and why changing one makes every number before and after it incomparable |
| The method surface | what you may change; this is also the module list your ablation-plan.tsv must cover |
| How to run one experiment | the exact command, where results land, and how to read the metric without pulling a training log into your context |
| The budget | how long one experiment may take before you kill it |
| Known breakages | traps already hit and diagnosed, so you do not rediscover them |
If STUDY.md is absent, stop and say so. Do not infer the study from the
code and start running: a brief you wrote yourself is not a brief, and the
frozen list in particular is not derivable from reading a repository — it is a
decision about what a number in this venue means.
A worked instance is in references/study-bound-to-disagree.md; copy it to the
study repository as STUDY.md if that is the study you are on.
What ships with it
13 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
- references/manuscript-readiness.md 8.1 KB
- references/study-bound-to-disagree.md 7.9 KB
- references/study-brief.md 6.2 KB
- references/validation.md 8.0 KB
- scripts/check_calibration.py 5.3 KB runs code
- scripts/check_design_validity.py 33 KB runs code
- scripts/check_kill_argument.py 7.8 KB runs code
- scripts/check_manuscript_readiness.py 44 KB runs code
- scripts/check_related_work.py 16 KB runs code
- scripts/check_reproduction.py 124 KB runs code
- scripts/plan_feasibility.py 8.1 KB runs code
- scripts/test_check_manuscript_readiness.py 44 KB runs code
- scripts/verdict.py 5.9 KB runs code
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.
- 1mo ago Changed · +7 lines f0c0ee9f02d8
- 1mo ago Changed · -7 lines 326868c21bae
- 1mo ago First seen · 606 lines · 63 tokens per session scan A 870197b027ea
evidence-preserving-research is a skill published in the GitHub repository AutoConference/AutoConference-skill (13 stars, last pushed 4d ago), licensed Apache-2.0. It adds 63 tokens to every session and 7,749 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 2026-09-24.
evidence-preserving-research compared with skills from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| evidence-preserving-researchAutoConference/AutoConference-skill | Skill | 13 | 4d ago | Apache-2.0 | A | 63 | Claude Code, Codex |
| instrument-data-to-allotropeanthropics/knowledge-work-plugins | Skill | 36k | today | Apache-2.0 | A | 123 | Claude Code |
| exploratory-data-analysisK-Dense-AI/scientific-agent-skills | Skill | 57k | 4d ago | MIT | A | 83 | Claude Code |
| matlabK-Dense-AI/scientific-agent-skills | Skill | 57k | 4d ago | MIT | A | 44 | Claude Code |
| phylogeneticsK-Dense-AI/scientific-agent-skills | Skill | 57k | 4d ago | MIT | A | 71 | Claude Code, Codex |
| research-engineerdavila7/claude-code-templates | Skill | 33k | yesterday | MIT | A | 43 | Claude Code, Codex |
| mapping-to-snomedmaziyarpanahi/openmed | Skill | 5.8k | 2d ago | Apache-2.0 | A | 205 | Claude Code |