Getting it into your agent
It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.
git clone --depth 1 https://github.com/romainsimon/skills-for-decision-makingnpx agentmods add skills/romainsimon/skills-for-decision-making/allocating-effortWrote 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/romainsimon/skills-for-decision-making/allocating-effort)<a href="https://agentmods.dev/skills/romainsimon/skills-for-decision-making/allocating-effort"><img src="https://agentmods.dev/badge/skills/romainsimon/skills-for-decision-making/allocating-effort/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/romainsimon/skills-for-decision-making/allocating-effort"><img src="https://agentmods.dev/badge/skills/romainsimon/skills-for-decision-making/allocating-effort.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 | $0.00093 | $0.01813 |
| Opus 5 | $0.00046 | $0.00907 |
| Sonnet 5 | $0.00019 | $0.00363 |
| Haiku 4.5 | $0.00009 | $0.00181 |
Grade A, and why
allocating-effort 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 12d 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 — 171 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Allocating effort
Deciding what to work on next is a bandit problem: several options, each with an unknown payoff rate, and every unit of effort spent on one is a unit not spent learning about the others. The two failure modes are symmetric and both are common - committing everything to the current leader before you know it is the leader, and spreading effort evenly long after you do.
Thompson sampling resolves both without tuning, and its allocation has a plain-English meaning: each option gets the share of effort equal to the probability it is the best one.
Workflow
- [ ] 1. Define the arms and what counts as a win
- [ ] 2. Count wins and losses honestly, including the zeros
- [ ] 3. Set priors where you genuinely know something
- [ ] 4. Run the allocation
- [ ] 5. Read the under-explored list before cutting anything
- [ ] 6. Commit the split for a fixed period, then recount
1. Define the arms and the win
An arm is anything you allocate to: a product, a channel, an ad creative, a landing page, a content format. The win must be:
- the same event for every arm - "signup" for one arm and "trial start" for another makes the comparison meaningless;
- as close to revenue as you can measure quickly - clicks are a fast proxy that frequently ranks arms in the opposite order to paid conversion;
- countable per unit of exposure - wins and losses, not a rate someone computed.
If arms have wildly different payoff sizes, this model is the wrong one: it compares rates, not values. Weight the wins by value first, or use the framing-decisions skill instead.
2. Count honestly, including the zeros
The arm with zero wins and fourteen tries goes in the table with wins: 0, losses: 14.
Leaving it out because "it obviously doesn't work" is the mistake this whole skill
exists to prevent: with a uniform prior that arm's posterior mean is 6%, not 0%, and
14 observations is nowhere near enough to distinguish 6% from 2%.
3. Set priors where you know something
What ships with it
5 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.
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.
- 12d ago First seen · 171 lines · 93 tokens per session scan A 1b33241fa93c
allocating-effort is a skill published in the GitHub repository romainsimon/skills-for-decision-making (11 stars, last pushed 1mo ago), licensed MIT. It adds 93 tokens to every session and 1,813 once invoked, about $0.0005 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.
Other skills, from other repositories
thinking-model-router
When unsure which thinking skill fits, map domain and problem type, then return NONE or one primary skill by default (at most three complementary).
thinking-systems
When behavior is emergent across components—fixes elsewhere break, loops/delays dominate—map boundary, stocks/flows, feedback, archetypes, then rank leverage.
architecture-aware-init
Selects architecture paradigm via research before scaffolding. Use when architecture is undecided and the choice needs justification and documentation.
thinking-five-whys-plus
When a fault is localized and the proximate cause is known but the systemic root is not, chain evidence-linked whys with a counterfactual stop and a countermeasure.
thinking-map-territory
When a claim, doc, test, metric, or assumption conflicts with observed behavior, stop theorizing from the map and verify the live code or data; let territory overrule.
thinking-theory-of-constraints
When throughput or latency is pipeline-limited, identify the single binding constraint and exploit, subordinate, elevate, then recheck—ignore non-constraints.