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 chan4lk/specclaw/plugin install specclawWrote 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/chan4lk/specclaw/propose)<a href="https://agentmods.dev/skills/chan4lk/specclaw/propose"><img src="https://agentmods.dev/badge/skills/chan4lk/specclaw/propose.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 | $0.00131 | $0.06811 |
| Opus 5 | $0.00066 | $0.03406 |
| Sonnet 5 | $0.00026 | $0.01362 |
| Haiku 4.5 | $0.00013 | $0.00681 |
Grade A, and why
propose 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 yesterday.
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 — 225 lines — stays where its author put it; the contents beside it link to each section on GitHub.
specclaw propose
First, run specclaw-ensure-init .specclaw — idempotently creates .specclaw/ if it doesn't exist (silent if already initialized; auto-inits using the current directory's basename as the project name).
Create a new proposal for a change.
If the user hasn't yet provided enough detail to draft the proposal (e.g. they just said "i have a proposal" with no specifics), ask once for the essentials inside this skill — what's the idea, what problem does it solve — then proceed to the steps below. Do not wait for a separate turn to invoke this skill.
- Target-foundation gate (brownfield rebuilds only — inert everywhere else). Run this before creating or renaming anything:
specclaw-bf-bootstrap foundation-check .specclaw
If applicable is false, continue to step 1b — there is no rebuild backlog, so this project is not a brownfield rebuild target and nothing here applies. Say nothing about it.
If the command does not run at all — command not found, any non-zero exit, or no JSON on stdout — you cannot read the gate, so fall back to the one signal that does not need it: if .specclaw/analysis/rebuild-backlog.md does not exist, continue to step 1b silently. No backlog means this is not a rebuild target, so the gate has nothing to protect and stopping an ordinary proposal because a brownfield-only binary is missing from an older install would be a pure false positive. If that file does exist, say the foundation gate could not be evaluated, name the command that failed, and stop — a brownfield project is exactly where this gate matters. The fail-closed rule at the bottom of this page governs a manifest this command read and distrusted; it does not govern a command that never ran.
If applicable is true and foundation_ready is false, stop. Do not offer or run the backfill, do not name the change, do not create the change directory, do not draft a proposal, do not run the dependency check. Relay the reason, then the remedy verbatim:
Target rebuild foundation has not been created. Run
/specclaw:bf-bootstrapfirst.
Why this gate is absolute. Without it, the first backlog item proposed inherits responsibility for inventing the application skeleton — and it inherits it invisibly, because nothing in the spec, the tasks or the verify report says "this item is also creating the app". That is how a screen-bearing item once shipped as a backend-only slice: the repo had no frontend, so the frontend was never anybody's task. A backlog item adds a capability to an application that already exists. If it does not exist yet, the answer is to create it deliberately, once, not to smuggle it into whichever item happened to be proposed first.
The one legitimate false positive is proposing an ordinary change inside the legacy repo, which also carries a rebuild-backlog.md. That is recorded once, explicitly, and attributably — never inferred, because there is no honest signal that distinguishes the two repos:
specclaw-bf-bootstrap not-applicable .specclaw \
--reason "<why this repo is not the rebuild target>" \
--declared-by "<name>, <YYYY-MM-DD>"
Both flags are required and the declaration is refused without them — same rule as a bypass's Chosen by. Ask the user for their name; do not supply one yourself.
When foundation_ready is true, continue silently. Mention the foundation only if the user asks.
1b. Name the change <NNN>-<slug> — e.g. 001-init-repo:
- Backfill offer — do this first, because a backfill changes the next number. Run
specclaw-renumber-changes .specclawwith no--apply: it is dry-run by default and renames nothing. If it prints a plan, unnumbered folders exist — show the user the fullold → newlist and ask once whether to apply it. On a yes, re-run with--apply; on a no, proceed with creating the new numbered change alone. No "already asked" flag is needed: the offer is conditioned on unnumbered folders existing, so it self-clears the moment a backfill runs. - Run
specclaw-next-change-number .specclawfor<NNN>, and slugify the user's idea (lowercase, hyphens, no spaces) for<slug>. Join them with a hyphen. Never format the number by hand —specclaw-next-change-numberowns that rule and is the only place it lives.
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.
- yesterday Changed · +154 lines c9c6dc3057f3
- 6d ago First seen · 71 lines · 131 tokens per session scan A 537a42febd98
propose is a skill published in the GitHub repository chan4lk/specclaw (12 stars, last pushed 2d ago), licensed MIT. It adds 131 tokens to every session and 6,811 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 2026-08-30.
Other skills, from other repositories
hns-lsel-curator
Local Self-Evolution Loop (LSEL) curator — the CLUSTER + drain engine for the GOOS-local PROPOSE→APPLY seam closure (SPEC-LSEL-LOCAL-EVOLUTION-001). Companion-offset drain of .moai/lessons-inbox.jsonl with a drain-side severity filter that drops the 65% Bash-timeout/sandbox noise, eventkey clustering with a frequency…
moai-foundation-core
Provides MoAI-ADK foundational principles including TRUST 5 quality framework, SPEC-First DDD methodology, delegation patterns, progressive disclosure, agent catalog reference, and token budget management (absorbed from moai-foundation-context). Use when referencing TRUST 5 gates, SPEC workflow, or context window…
moai-workflow-worktree
Git worktree management for parallel SPEC development with isolated workspaces, automatic branch registration, and seamless MoAI-ADK integration. Use when setting up parallel development environments.
moai-design-tools
Design tool integration specialist covering Figma MCP, Pencil renderer, and Pencil-to-code export. Use when fetching design context from Figma, rendering Pencil designs, or exporting to React/Tailwind code.
hns-workflow-ci-loop
Unified CI watch + auto-fix loop skill. Polls gh pr checks after /moai sync PR creation, classifies required vs auxiliary failures, attempts safe automated patches (max 3 iterations), and escalates semantic failures to the user. Use for CI loop workflow — NOT for general loop iteration patterns (see…
moai-kanban-foreman
One unattended kanban foreman iteration: watch the backlog queue, dispatch the next operator-picked card to an isolated worker, collect completion evidence on read (not on claims), and report. This is the body the project's loop.md driver invokes each iteration of a bare /loop; it can also be invoked directly to test…