AgentStack
Browse Sign in
Browse Why AgentStack Sell Docs
Sign in
SKILL verified Apache-2.0 Self-run

Builder

skill-mifunedev-agro-builder · by mifunedev

|

— No reviews yet
0 installs
10 views
0.0% view→install

Install

$ agentstack add skill-mifunedev-agro-builder

✓ scanned · ✓ verified, works with Claude Code, Cursor, and more.

Security review

✓ Passed

No issues found. Passed automated security review. · v0.1.0 How review works →

  • ✓ Prompt-injection patterns
  • ✓ Secret / credential exfiltration
  • ✓ Dangerous shell & filesystem operations
  • ✓ Untrusted network calls
  • ✓ Known-malicious package signatures

What it can access

  • ✓ Network access No
  • ✓ Filesystem access No
  • ✓ Shell / process execution No
  • ✓ Environment & secrets No
  • ✓ Dynamic code execution No

From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.

View the full security report →

Verified badge

Passed review? Show it. Paste this badge into your README, it links to the public security report.

AgentStack Verified badge Links to your public security report.
[![AgentStack Verified](https://agentstack.voostack.com/badges/verified.svg)](https://agentstack.voostack.com/security/report/skill-mifunedev-agro-builder)

Reliability & compatibility

✓ Security review passed
0 installs to date
— no reviews yet
● 16d ago

Declared compatibility

Claude CodeClaude Desktop

Compatibility is declared by the source manifest. End-to-end runtime verification is coming, see below.

Preview Execution monitoring

We're building live execution health for every listing: tool-call success rate, median latency, uptime, and last-checked timestamps, measured, not self-reported. It isn't live yet, so we don't show numbers we can't stand behind.

How agent discovery & health will work →
Are you the author of Builder? Claim this listing to set pricing, connect Stripe payouts, and keep 70% of every sale.
Sign up to claim

About

Builder

Build one artifact through the matching type reference. Run inline and inherit the session model; do not fork or override the model unless the artifact being authored has an independently justified need.

Arguments received: $ARGUMENTS

Dispatch

  1. Treat the first whitespace-delimited argument as TYPE and the remainder as

the artifact name, request, or path.

  1. Accept exactly these types:

| Type | Read and follow | |------|-----------------| | skill | references/skill.md | | command | references/command.md | | rule | references/rule.md |

  1. If TYPE is missing or unknown, or the remaining request is empty or only

whitespace, print the following and stop without reading type references or modifying files:

``text Usage: /builder ``

agent is not an artifact type. A reusable role, procedure, or specialist judgment is authored as a skill; a bounded isolated worker context is an execution choice made by /delegate, not a repository artifact.

  1. Read the selected reference completely, then execute its protocol against the

remainder of $ARGUMENTS. The selected reference is authoritative for artifact shape and type-specific validation.

Shared protocol

Apply these steps for every valid type before the selected reference's type-specific steps.

1. Discover local authority

  • Find and read applicable AGENTS.md and CLAUDE.md files from repository root

through the target directory. More local instructions win; in one directory, AGENTS.md is canonical.

  • Identify the source-of-truth artifact directory. In Open Harness and equipped

projects, edit .agro/skills/; provider directories such as .claude/, .codex/, and .pi/ are generated or symlinked exposure surfaces.

  • Outside an Open Harness layout, follow the target project's documented canonical

path rather than creating .agro/ speculatively.

  • Inspect two or three nearby artifacts of the same type. Reuse their naming,

frontmatter, structure, tone, and validation conventions.

  • Search for an existing artifact with the same purpose. Prefer a focused update or

explicit consolidation over a near-duplicate.

  • Consult compiled harness patterns before proposing a change. Run

/wiki query --patterns and read what it returns: each page records a failure mode, its root cause, and a workaround this harness already paid for. Cite the motivating [[pattern-...]] slugs in the report.

  • Read .agro/evals/decisions/skill-impact.md for prior proposals against the same

target. Do not re-propose a change recorded there as REJECTED unless new evidence contradicts the recorded validation; when you do, cite the prior record id.

2. Define the contract

Before editing, state internally:

  • the artifact's one-sentence purpose and concrete triggers;
  • who invokes or consumes it;
  • what is in scope, out of scope, and considered done;
  • the minimum tools, context, side effects, and supporting resources required;
  • which behavior is repository-specific and must be grounded in inspected files.

Ask a question only when unresolved ambiguity would materially change the artifact or create unsafe side effects. Otherwise use the request and repository evidence.

3. Author narrowly

  • Use lowercase kebab-case names and one artifact per coherent concern.
  • Put matching and trigger information in frontmatter, not only in the body.
  • Use imperative, operational language. Remove generic expertise prose that does

not change behavior.

  • Prefer the least privilege and smallest context footprint that completes the job.
  • Cite real local paths and commands only after verifying them.
  • Do not modify unrelated files, generated provider mirrors, or user work in the

working tree.

4. Validate and report

  • Validate frontmatter delimiters and required fields without assuming optional YAML

libraries are installed.

  • Check every referenced path, invocation, tool, and supporting file.
  • Enforce the selected reference's size, safety, and semantic checks.
  • When .agro/scripts/link-providers.sh exists and canonical .agro/ primitives were

changed, run bash .agro/scripts/link-providers.sh --check.

  • Run git diff --check when inside a Git worktree.
  • Report the files created, updated, or removed; the resulting invocation or loading

behavior; key design choices; and validation evidence. Never claim a check ran if it did not.

  • Append a PROPOSED record to .agro/evals/decisions/skill-impact.md when a skill

edit lands: the next SI-nnnn id, the one-sentence proposal, the single target artifact, the motivating pattern slugs, and the unified diff scoped to that target path. motivating patterns: none (direct request) is a legitimate value — record it rather than inventing a pattern to cite. Stage the ledger (.agro/evals/decisions/ is tracked, so a plain git add) and report the allocated id. Never edit an existing record.

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

Install and usage instructions live in the source repository linked above.

Reviews

No reviews yet, be the first.

Versions

  • v0.1.0 Imported from the upstream source.