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/brd-intake)<a href="https://agentmods.dev/commands/ihudak/ihudak-claude-plugins/brd-intake"><img src="https://agentmods.dev/badge/commands/ihudak/ihudak-claude-plugins/brd-intake/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/brd-intake"><img src="https://agentmods.dev/badge/commands/ihudak/ihudak-claude-plugins/brd-intake.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 | $0.00251 | $0.32878 |
| Opus 5.5 | $0.00100 | $0.13151 |
| Sonnet 5.5 | $0.00050 | $0.06576 |
| Haiku 4.5 | $0.00025 | $0.03288 |
Grade A, and why
brd-intake 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 — 1,565 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Intake the customer-supplied business requirements document: $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: ${CLAUDE_PLUGIN_ROOT} resolves to this plugin, which does not carry it.
/brd-intake is the entry point of the BRD-to-PRD flow (PM phase) — the first of the
route's commands, which between them turn a long, often internally
contradictory customer BRD into requirements a PRD can be built from. It copies the customer's
source into the specs repo verbatim and immutably, extracts a [BR#n] requirement inventory
via the brd-reader agent, classifies the document's defects with a human rather than on the
agent's say-so alone, and writes a coverage ledger in which every requirement starts life
unallocated — the state /brd-split, the route's allocation step, cannot complete past until each
row has been given a fate.
Usage: /brd-intake <BRD-KEY> @<brd-file> [--sort-existing <dir>] [--no-docs] [--docs <path>]
Phase 0 — Resolve inputs
<BRD-KEY>(mandatory). Parse the first token that is neither a flag, nor a flag's value, nor the@<brd-file>argument —--sort-existingand--docseach consume the token after them (step 4), and a value skipped as "non-flag" would be read as the key; a token opening with@is step 2's source wherever it stands, so an operator who types the path first is not told the key is missing; validate it withkey-valid(workflows-core:addressing§1'skey-valid— shape only, never checked against a tracker). If absent or invalid, stop gracefully:BRD_INTAKE_NEEDS_KEY: /brd-intake needs a BRD key (shape ^[A-Z][A-Z0-9_]*(-\d+)+$, e.g. ACME-001) — pick a short stable identifier for this business requirements document, then re-run '/product-workflows:brd-intake <KEY> @<brd-file>'.@<brd-file>(mandatory). The customer's source file argument. If absent, stop:BRD_INTAKE_NEEDS_SOURCE: /brd-intake needs the customer's source as an @-argument — re-run '/product-workflows:brd-intake <KEY> @<path-to-brd>'.- Reject a PDF — do not convert it. If the resolved source does not end in
.md/.markdown(a PDF, a Word document, a slide deck, any non-markdown source), stop:BRD_INTAKE_NEEDS_MARKDOWN: the source must be markdown — convert it first, and check the conversion. It becomes the immutable record every [BR#n] anchors into.The reason this is a hard stop rather than a best-effort conversion: once intaken,brd/source/is never edited again (${CLAUDE_PLUGIN_ROOT}/references/brd-format.md§1), and every[BR#n]locates itself inside that text bysource_anchor. An unchecked machine conversion — a dropped clause, a misplaced heading, a table read out of row order — would silently become the record of what the customer asked for, with no later step positioned to catch it. Converting is the operator's own step, done where they can eyeball the result against the original before handing it back to this command. - Optional flags.
--sort-existing <dir>— if present, validate<dir>exists and carry it forward to Phase 6.--docs <path>— points documentation grounding at that root for this run instead of${DOCS_PATH:-/workspace/docs}; strip the flag and its value together before any remaining-argument classification, or the path is read as part of the address. Declared for every consumer byworkflows-core:docs-groundingProcedure step 1 (Flags first), which resolves it; this command only has to recognise it and pass the invocation through.--no-docs— boolean; turns documentation grounding off for this run, carried to Phase 1'sresolve-docs-groundingcall. None of the three changes anything else about Phase 0: the BRD source is still required and still gated by step 3. $SPECS_PATH(required). If unset, stop namingSPECS_PATH, per theRequired path environment variable unsetrule inworkflows-core:escalation-rules:choices: ["Set SPECS_PATH (enter the path)", "Cancel"]- Specs-repo preflight. Invoke
Skill(skill: "workflows-core:reference", args: "specs-repo-git specs-preflight")and execute itsspecs-preflightentry point (§3) inline. Prompt-free and silent when the specs repo is clean and on its default branch. If a guard fires, emit its §5 notice; if it returnsspecs_git: blocked(§3.3 G0), carry that flag for the whole run — the terminalcommit-artifactsstep skips on it. - Resolve or derive the BRD folder via
resolve-address <BRD-KEY>(Skill(skill: "workflows-core:reference", args: "addressing resolve-address"), §3). Found → an existing folder, and where it is a root BRD this invocation is a re-run over it; use it. Only a root is: the slice test below and the level test after it refuse every other folder. A slice is never a legitimate target here — it has no source document of its own to intake, and its inventory and ledger are created by the parent's/brd-split(${CLAUDE_PLUGIN_ROOT}/references/brd-format.md§2.1,${CLAUDE_PLUGIN_ROOT}/references/coverage-ledger-format.md§3) — so a resolved folder whosebrd-link.mdcarries aparent:field is a mis-keyed invocation, and the run stops on it. Take this refusal here, before the re-run confirmation below is composed, and note that it keys onparent:rather than on rows because neither of the two states a slice can be in is safe. A slice that claims any row holds an inventory with those rows in it,/brd-splithaving copied the parent's in, so there that confirmation would otherwise fire and be the only array the operator sees on this path — an array whose keeps-and-changes list and whose disposition warning are both written for a source-owning root, and whose Re-run over this folder answer sends Phase 2 on to writebrd/source/into a folder that holds none, Phase 3 to renumber rows §2.1 says are never re-extracted, and Phase 5 to replace everycovered-herethe allocate-only walk recorded withunallocated. Nothing short of git recovers that. And on a standing empty child — one kept withclaims: []against a recordedreason:(coverage-ledger-format.md§3), whose inventory holds its header and no row, which this step's own test reads as no prior inventory at all — that confirmation is skipped silently, so Phase 2 would write into the slice with no consent asked anywhere. Stop, on a slice:BRD_INTAKE_SLICE: <BRD-KEY> resolves to a slice of <the parent: field of its own brd-link.md>, and a slice has no source document to intake — the source and the records beside it under brd/ are the parent's, one hop, and the slice's own brd/brd-inventory.md and coverage-ledger.md are created by '/product-workflows:brd-split' on the parent. Re-intake the parent with '/product-workflows:brd-intake <PARENT-KEY> <the @<brd-file> this run was given>'; that run leaves every row of the parent's ledger unallocated, so re-take the dispositions with the root walk it offers you when it finishes, '/product-workflows:brd-split <PARENT-KEY> "<how to cut it>"', and only then re-allocate a slice the walk left a row unallocated in, with '/product-workflows:prd-ground <SLICE-KEY>' and then '/product-workflows:brd-split <SLICE-KEY>'.Four of the six slots are substituted and two stay literal, and the text says which.<PARENT-KEY>, and the slot naming theparent:field in the stop's first sentence, are both thatparent:field as it reads, never a key parsed out of the folder's name (CLAUDE.md, Resolve an identifier against a known set);<BRD-KEY>is the address typed; and the source is the@<brd-file>this invocation was given, substituted exactly as typed — the run holds it, and it is telling the operator to re-type it against a different key."<how to cut it>"stays literal, as Phase 8 prints its own copy, because the slicing instruction is the operator's to write; and<SLICE-KEY>stays literal because the root walk that would carve or confirm that slice has not run yet. Thebrd/half is stated by what is the parent's rather than by the directory: a slice does hold abrd/of its own, holding exactlybrd/brd-inventory.md(brd-format.md§2.1) — which is why naming the whole directory the parent's would contradict the clause after it.
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 · +37 lines 4098c6236065
- 1mo ago Changed · +957 lines · +64 tokens per session 34aafe178ddb
- 1mo ago Changed · +63 lines · +36 tokens per session 8ea72722606c
- 2mo ago Changed · +33 lines 2f678f7b2d20
- 2mo ago Changed · -2 lines · +1 tokens per session 96ae5d0d18c6
- 2mo ago Changed · +88 lines fc892cc3a572
- 2mo ago First seen · 389 lines · 150 tokens per session scan A 2d300733093f
brd-intake is a command published in the GitHub repository ihudak/ihudak-claude-plugins (2 stars, last pushed 1mo ago), licensed MIT. It adds 251 tokens to every session and 32,878 once invoked, about $0.0010 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 2026-08-31.
brd-intake compared with commands from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| brd-intakeihudak/ihudak-claude-plugins | Command | 2 | 1mo ago | MIT | A | 251 | 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 |