grasp

grasp is a skill for Claude Code from hajibabaie/combinatorial-optimization-skills. It costs 121 tokens per session (9,124 once invoked), scanned A, original, MIT.

GRASP, or Greedy Randomized Adaptive Search Procedure, is a repeated two-stage search method. It builds a solution by making partly random greedy choices, then improves it with local search, and keeps the best result across many starts.

In plain words
What is it for?
Use it for combinatorial problems where solutions can be built one element at a time and then improved by small changes, such as assignments, routes, or selections.
Why use it?
A purely greedy method can repeatedly make the same poor early choice, while a single local search can settle on a nearby answer. GRASP adds controlled variation and repeated improvement to explore more alternatives.

Skill for Claude Code

Written for Claude Code: shipped in a Claude Code plugin. Also seen: positional $N argument.

Part of the combinatorial-optimization plugin — 76 skills shipped together

Good fit Use it for combinatorial problems where solutions can be built one element at a time and then improved by small changes, such as assignments, routes, or selections.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/hajibabaie/combinatorial-optimization-skills/grasp
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.

Any agent
npx skills add hajibabaie/combinatorial-optimization-skills --skill grasp
Clone the repo
git clone --depth 1 https://github.com/hajibabaie/combinatorial-optimization-skills

Made for: Claude Code.

Or install combinatorial-optimization, the plugin that ships this one along with the rest of its 76 skills.

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 grasp

README.md
[![agentmods](https://agentmods.dev/badge/skills/hajibabaie/combinatorial-optimization-skills/grasp.svg)](https://agentmods.dev/skills/hajibabaie/combinatorial-optimization-skills/grasp)
Your own site
<a href="https://agentmods.dev/skills/hajibabaie/combinatorial-optimization-skills/grasp"><img src="https://agentmods.dev/badge/skills/hajibabaie/combinatorial-optimization-skills/grasp.svg" alt="Measured on agentmods" height="20"></a>
Per session 121 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 9,124 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
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 5.1 $0.00121 $0.09124
Opus 5 $0.00060 $0.04562
Sonnet 5 $0.00024 $0.01825
Haiku 4.5 $0.00012 $0.00912

Measured 8d ago against content hash c329da72cee0, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-08, from the pricing page.

Security

Grade A, and why

grasp 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.

skills/grasp/SKILL.md · 779 lines

How it starts

The opening of the file, as written. The whole thing — 779 lines — stays where its author put it; the contents beside it link to each section on GitHub.

GRASP — Greedy Randomized Adaptive Search Procedures

You are an expert in metaheuristic optimization, specializing in GRASP and multi-start methods. This skill covers greedy randomized construction with restricted candidate lists (RCL), the construction/local-search iteration, alpha calibration, reactive GRASP, bias functions, and hybridization with path relinking. Use the framework below to decide whether GRASP fits a problem, implement it correctly in vectorized numpy, and report results that withstand reviewer scrutiny.

Initial Assessment

Establish the following before proposing or writing any GRASP code:

  • Greedy structure. Identify the element-by-element construction: what is an "element" (a column, an assignment, an edge), and what incremental greedy score does adding it have? GRASP requires a cheap, meaningful greedy function. If none exists, reconsider the method.
  • Problem class and objective. Minimization or maximization, single objective, and which constraints the construction must respect at every partial step.
  • Hard vs soft constraints. Hard constraints shape the candidate set during construction; soft constraints belong in the objective or in penalties.
  • Instance size. Number of elements per solution, candidates per construction step, and whether greedy scores can be updated incrementally instead of recomputed.
  • Local search. Which neighborhood improves constructed solutions, and whether O(1)/O(n) delta evaluation exists. GRASP without a competent improvement phase is just semi-greedy sampling and usually underperforms.
  • Evaluation cost. Time per construction and per local-search descent determines how many restarts fit the budget.
  • Exact-vs-heuristic need. If instances are small enough for a MIP solver to close the gap in the available time, solve exactly and use GRASP only for warm starts.
  • Time budget and stopping rule. Wall-clock limit, iteration limit, or stagnation-based stop; restarts are independent, so the budget maps linearly to iterations.
  • Quality requirement. Target gap to best-known solutions or bounds; whether a single good solution or a distribution over seeds is required.
  • Data format. How instances arrive (matrix files, benchmark formats, generated) and what an independent feasibility check looks like.
  • Memory features in scope. Plain GRASP is memoryless; decide up front whether reactive alpha, hashing of duplicates, or path relinking is part of the deliverable.
  • Parallel resources. Independent restarts parallelize with near-linear speedup; count available cores.
  • Reproducibility. One np.random.default_rng(seed) per run; record seeds in all result tables.

Read the full file on GitHub · 779 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. 8d ago First seen · 779 lines · 121 tokens per session scan A c329da72cee0

Subscribe to this mod's changes

grasp is a skill published in the GitHub repository hajibabaie/combinatorial-optimization-skills (7 stars, last pushed 2mo ago), licensed MIT. It adds 121 tokens to every session and 9,124 once invoked, about $0.0006 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.

Related

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.

oliver-kriska/claude-elixir-phoenix · 58 tokens

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.

oliver-kriska/claude-elixir-phoenix · 60 tokens

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.

oliver-kriska/claude-elixir-phoenix · 40 tokens

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.

oliver-kriska/claude-elixir-phoenix · 48 tokens

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.

oliver-kriska/claude-elixir-phoenix · 39 tokens

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.

oliver-kriska/claude-elixir-phoenix · 46 tokens