Getting it into your agent
This one installs as part of its plugin. Adding the marketplace and installing the plugin brings it with everything else the plugin ships.
/plugin marketplace add ihudak/ihudak-claude-plugins/plugin install product-workflowsWrote 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/commands/ihudak/ihudak-claude-plugins/prd-proposal)<a href="https://agentmods.dev/commands/ihudak/ihudak-claude-plugins/prd-proposal"><img src="https://agentmods.dev/badge/commands/ihudak/ihudak-claude-plugins/prd-proposal/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/commands/ihudak/ihudak-claude-plugins/prd-proposal"><img src="https://agentmods.dev/badge/commands/ihudak/ihudak-claude-plugins/prd-proposal.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 | $0.00195 | $0.14284 |
| Opus 5.5 | $0.00078 | $0.05714 |
| Sonnet 5.5 | $0.00039 | $0.02857 |
| Haiku 4.5 | $0.00019 | $0.01428 |
Grade A, and why
prd-proposal 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 — 759 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Author a customer-facing effort proposal for the resolved PRD folder: $ARGUMENTS
Core references. A citation of the form workflows-core:<name> names a shared reference in the workflows-core plugin. Load it with Skill(skill: "workflows-core:reference", args: "<name>") — never by path: ${CLAUDE_PLUGIN_ROOT} resolves to this plugin, which does not carry it.
/prd-proposal is the effort-proposal step (PM phase) — it takes one PRD- folder, grades how much
evidence stands behind it, and writes the document a vendor sends a customer: what the requirement set
will take in human delivery hours, by work package and by role, as a range whose width is computed
from per-package confidence and whose every cost driver resolves to a record on disk. The canonical
shape of both artifacts — their section sets, the two identifier namespaces, the readiness tiers, the
confidence grades and the closed evidence set — is
${CLAUDE_PLUGIN_ROOT}/references/proposal-format.md. This command executes that file; it never
restates it.
Usage: /prd-proposal <ADDRESS> [--no-brief] [--profile] [--baseline <path>] [--redo]
<ADDRESS> resolves through workflows-core:addressing §3 resolve-address, taken after
$SPECS_PATH is settled — a resolution taken before the variable is known returns absent for a
folder that exists, which is the ordering /create-prd and /create-ard state for the same reason.
This command gates nothing on the build ladder and nothing on it waits. It gates its own input
and nothing beyond it. No command of the build ladder reads proposal.md — /create-ard,
/specify, /epics, /dev-workflows:design, /dev-workflows:implement and /dev-workflows:ready
each resolve the same folder and neither know nor care whether it holds one — so nothing there is
waiting on this run and no readiness tier withholds permission to build. The one command that reads
another folder's proposal as a proposal is the sibling umbrella
/product-workflows:brd-proposal, which is a second proposal rather than a phase of the build;
every other targeted read is an own-folder one — a later run of this command, which opens the
proposal.md it left in this same folder as §8's stability anchor (Phase 0's revision note, Phase 6
step 6), and proposal-reviewer inside the run that wrote it (Phase 9).
/product-workflows:brd-reconcile's stale cross-reference sweep also reads any proposal under the
parent BRD, as ordinary prose, and never edits one — a stale reference in it goes to what still
needs a human, fixed by re-running this command. Running this is optional at every tier, in the
same sense /prd-ground is optional and ungated on the idea route. A proposal is a document a
vendor sends a customer — not a phase, not a prerequisite, and never a reason implementation cannot
start.
/prd-proposal is not BRD-route-only. A PRD- folder is a PRD- folder. An idea-route PRD
carrying no brd-link.md, no decisions.md and no grounding/ estimates fine — it grades at tier 1
(${CLAUDE_PLUGIN_ROOT}/references/proposal-format.md §5), and the document says outright that its
drivers are unknown.
No money for human hours, at any tier, under any flag — no rate, no currency symbol, no monetary
total, in either artifact or in the profile
(${CLAUDE_PLUGIN_ROOT}/references/proposal-format.md §1). The USD figures this run emits at the end
are model spend and are a different quantity.
Both artifacts are prose, and prose is never hard-wrapped — one unbroken line per paragraph,
per workflows-core:prose-formatting, so a straight copy-paste into a review tool or a message needs
no cleanup. This governs what the run writes; this command file itself stays wrapped like its
neighbours.
There is no writer agent, and that is deliberate. The proposal is authored inline by this command,
as prd.md, ard.md and specification.md are: the authoring is the command's purpose, and a
handoff to a writer would only add a place for the resolved data set to be lost between derivation and
rendering.
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 · +71 lines e1452f9ca49d
- 1mo ago Changed · +5 lines 91a9d81e0fb5
- 1mo ago Changed · +12 lines a2a47e334424
- 2mo ago First seen · 671 lines · 195 tokens per session scan A 5aa0e129d841
prd-proposal is a command published in the GitHub repository ihudak/ihudak-claude-plugins (2 stars, last pushed 1mo ago), licensed MIT. It adds 195 tokens to every session and 14,284 once invoked, about $0.0008 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 2026-09-10.
prd-proposal compared with commands from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| prd-proposalihudak/ihudak-claude-plugins | Command | 2 | 1mo ago | MIT | A | 195 | Claude Code |
| battlecardphuryn/pm-skills | Command | 28k | 2d ago | MIT | A | 19 | Claude Code |
| memorizelucasrosati/claude-code-memory-setup | Command | 1.1k | 21d ago | MIT | A | 0 | Claude Code |
| watchMarcel-Bich/marcel-bich-claude-marketplace | Command | 12 | 2d ago | MIT | A | 11 | Claude Code |
| opsfrankxai/agentic-creator-os | Command | 12 | 2d ago | Apache-2.0 | A | 35 | Claude Code |
| value-prop-statementsstefanoskarakasis/Product-Marketing-Skills | Command | 11 | 2d ago | MIT | A | 16 | Claude Code |
| commitjoaquimscosta/arkhe-claude-plugins | Command | 22 | 26d ago | MIT | A | 11 | Claude Code |