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 agentmods add skills/melodic-software/claude-code-plugins/goalnpx skills add melodic-software/claude-code-plugins --skill goalgit clone --depth 1 https://github.com/melodic-software/claude-code-pluginsWrote 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/melodic-software/claude-code-plugins/goal)<a href="https://agentmods.dev/skills/melodic-software/claude-code-plugins/goal"><img src="https://agentmods.dev/badge/skills/melodic-software/claude-code-plugins/goal.svg" alt="Measured on agentmods" 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 | $0.00178 | $0.01737 |
| Opus 5 | $0.00089 | $0.00869 |
| Sonnet 5 | $0.00036 | $0.00347 |
| Haiku 4.5 | $0.00018 | $0.00174 |
Grade A, and why
goal 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 2d 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 — 134 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Purpose
Answers "what would count as done, and is it reachable at all?"
The failure this prevents: the source run behind this plugin set a goal of p50 <= 250 ms for a hook on a host that charged 0.3-2.8 s for a single irreducible process spawn. The goal was unreachable by any code change, and that was discovered at the end, after the work. Knowing it up front would have reframed the entire task from "make it fast" to "remove the spawn or accept the floor".
Human-gated, always
This phase requires the user. /performance:snapshot and /performance:verify may run unattended;
this one may not, under any autonomy setting.
The reason is specific: computing the floor routinely produces a verdict the user has to act on ("this target is unreachable"), and choosing between a reframed goal, a different target, and accepting the floor is a judgment about what the work is for. An agent resolving that on the user's behalf converts a surfaced constraint back into a silent one.
If the user is unavailable, stop and say what is blocked. Do not pick a target and proceed.
What a goal must contain
1. The metric, and the exact command that produces it
Not "latency". The literal command, its arguments, and the field of its output that is the number. A metric nobody can re-run is not a metric.
Name the drift-immune counter alongside it (spawns, syscalls, queries, allocations, round trips) and rank the counter above the duration. On a host that cannot support a wall-clock claim, the counter is what survives; in the source run the durable result was a spawn census of 4 -> 1, and the milliseconds were not reproducible by an independent verifier on the same machine an hour later.
2. The floor, computed before any work
The irreducible cost this target cannot go below whatever the code does. Compute it by measuring the cheapest possible version of the operation: the empty hook, the no-op spawn, the single round trip, the query returning one row.
lib/spawn_noise.py's spawn_probe() gives the process-spawn floor for this host directly.
What ships with it
1 file 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.
- 2d ago First seen · 134 lines · 178 tokens per session scan A 2492ecce5496
goal is a skill published in the GitHub repository melodic-software/claude-code-plugins (15 stars, last pushed today), licensed MIT. It adds 178 tokens to every session and 1,737 once invoked, about $0.0009 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
parallel-orchestrator
Manage parallel Claude Code workstreams using git worktrees. Use when: splitting large tasks across multiple workers, coordinating parallel development, monitoring worker progress, integrating completed work, analyzing work item documents (code reviews, issue lists). Triggers: parallel, orchestrator, worktrees…
parallel-worker
Execute focused implementation tasks in a parallel workflow. Use when: working on assigned files in a worktree, making checkpoint commits, signaling dependencies or blockers, completing orchestrator-assigned tasks. Triggers: worker, checkpoint, worktree, assigned scope, commit prefix, parallel task.
solve-constraint-puzzle
For constraint satisfaction: Sudoku, scheduling, N-queens, logic puzzles, SAT-like problems, assignment problems. Uses propagate-then-search pattern.
use-class-for-state
For complex state: encapsulate multiple interacting variables, stateful algorithms, backtracking search state.
build-priority-queue
For ordered processing: A search, Dijkstra, event simulation, task scheduling. Efficient min/max extraction with heap-based queue.
catch-expected-errors
For iteration with errors: catch exceptions during exploration, skip invalid cases, continue to next attempt.