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 hajibabaie/combinatorial-optimization-skills --skill warm-starts-and-initial-solutionsgit clone --depth 1 https://github.com/hajibabaie/combinatorial-optimization-skillsWrote 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/hajibabaie/combinatorial-optimization-skills/warm-starts-and-initial-solutions)<a href="https://agentmods.dev/skills/hajibabaie/combinatorial-optimization-skills/warm-starts-and-initial-solutions"><img src="https://agentmods.dev/badge/skills/hajibabaie/combinatorial-optimization-skills/warm-starts-and-initial-solutions/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/hajibabaie/combinatorial-optimization-skills/warm-starts-and-initial-solutions"><img src="https://agentmods.dev/badge/skills/hajibabaie/combinatorial-optimization-skills/warm-starts-and-initial-solutions.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 | $0.00132 | $0.11594 |
| Opus 5 | $0.00066 | $0.05797 |
| Sonnet 5 | $0.00026 | $0.02319 |
| Haiku 4.5 | $0.00013 | $0.01159 |
Grade A, and why
warm-starts-and-initial-solutions 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 8d 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 — 705 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Warm Starts and Initial Solutions
You are an expert in initial-solution engineering for combinatorial optimization: building good starting solutions cheaply, injecting them into exact solvers (MIP starts, variable hints, partial fixing, basis reuse), and moving solutions the other way — from exact solvers into metaheuristic populations. Use the framework below to decide which mechanism fits, to run the injection protocol end to end, and to measure whether the warm start actually paid for itself. This is a methodology skill: the protocols matter as much as the code.
Initial Assessment
Establish these facts before touching any Start attribute. Most failed warm starts trace back to a skipped item here.
- Direction of transfer. Heuristic solution into an exact solver, exact solution into a metaheuristic, or solver-to-solver across a reoptimization sequence? Each direction has its own mechanism and its own failure modes.
- Source of the start. A construction heuristic, a previous solve on slightly different data, a rounded LP relaxation, or a human-made plan? The source determines how much of it you can trust and whether it is feasible for the current model.
- Complete or partial values. Does the candidate cover every variable family, or only the "important" ones (e.g., the binaries)? Partial starts and complete starts go through different mechanisms.
- Feasibility for THIS model. A solution feasible for last week's data, or for a model variant without one constraint family, is not a MIP start — it is a hint at best. Verify against the current model before injecting.
- Where the pain is. Time to first feasible solution, incumbent quality at the time limit, or proving optimality? Warm starts improve only the primal side; if the dual bound is the bottleneck, a warm start will not close the gap.
- Solver and API. Gurobi (
Start,VarHintVal,NumStart, basis attributes), OR-Tools CP-SAT (AddHint), HiGHS, PuLP/Pyomo passthrough? The mechanism names and semantics differ. - Model identity across runs. Is the model object kept alive, or rebuilt from scratch each run? Rebuilt models need name-based variable mapping; object references die with the old model.
- Is this a one-off solve or a sequence? Rolling horizons, branch-and-price iterations, and parameter sweeps re-solve near-identical models; there the start should come from the previous solution, not from a constructor.
- Cost of constructing the start. A constructor that takes 30 s to save 10 s of solver time is a net loss. Budget construction time against expected savings.
- Metaheuristic side: population size and seed fraction. How many seeds are available, how diverse are they, and what fraction of the population should they occupy? Seeding 100% of a population with near-identical elites kills diversity on arrival.
- Measurement plan. Same time limit, with/without comparison, several seeds for randomized components. Decide the protocol before running, or the result will be an anecdote.
- Tolerances.
IntFeasToland feasibility tolerances decide whether a numerically noisy start is accepted. Values like 0.9999999 must be rounded before injection, not after rejection.
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.
- 8d ago First seen · 705 lines · 132 tokens per session scan A c8a67de93877
warm-starts-and-initial-solutions is a skill published in the GitHub repository hajibabaie/combinatorial-optimization-skills (7 stars, last pushed 3mo ago), licensed MIT. It adds 132 tokens to every session and 11,594 once invoked, about $0.0007 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 2026-09-03.
Other skills, from other repositories
phx-deps-audit
Audit Hex deps for supply-chain security risk — bidi chars, compile-time exec, maintainer changes, typosquats, CVEs. Use after mix deps.update, when checking if a package upgrade is safe, or reviewing mix.lock PR diffs.
release
CONTRIBUTOR TOOL - Cut a plugin release: bump plugin.json version, finalize CHANGELOG, update README if needed, gate on make ci, commit, tag vX.Y.Z, and create the GitHub release. Use when shipping a new plugin version. NOT distributed.
session-deep-dive
Deep qualitative analysis of high-signal sessions. Spawns subagents with v2 template, synthesizes patterns, compares against known findings. Use after /session-scan.
catchup
Summarize and review what changed while you were away. Use after a weekend, vacation, or flight to check missed PRs, git commits, Linear tickets, and meetings — one prioritized brief, not a firehose.
brainstorm
Brainstorm Elixir/Phoenix features — explore ideas, compare approaches, gather requirements. Use when vague idea, not sure how to approach, or want to discuss before plan.
learn-from-fix
Capture Elixir/Ecto/LiveView lessons and Hex API rules. Use after corrections or when asked to document learning, record a lesson, prevent a fixed mistake, or remember package guidance with --library.