init-verify

init-verify is a skill for Claude Code from cryndoc/polisade-orchestrator. It costs 91 tokens per session (1,328 once invoked), scanned A, original, Apache-2.0.

A structural check for confirming that project initialization created the required files and valid metadata. Project initialization is the first setup step that creates the project's state, context, and configuration files.

In plain words
What is it for?
Use it after running the project's initialization command to verify the state JSON, version fields, context file, and environment template.
Why use it?
It catches setup output that looks plausible but is missing required fields, uses incorrect version information, or contains placeholder values. It checks the project files directly rather than rebuilding them from memory.

Skill for Claude Code

Written for Claude Code: Claude Code plugin machinery. Also seen: mentions CLAUDE.md.

Part of the polisade plugin — 26 skills, 2 commands shipped together

Install

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.

agentmods
npx agentmods add skills/cryndoc/polisade-orchestrator/init-verify
Any agent
npx skills add cryndoc/polisade-orchestrator --skill init-verify
Clone the repo
git clone --depth 1 https://github.com/cryndoc/polisade-orchestrator

Made for: Claude Code.

Or install polisade, the plugin that ships this one along with the rest of its 26 skills, 2 commands.

Wrote 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.

agentmods badge for init-verify

README.md
[![agentmods](https://agentmods.dev/badge/skills/cryndoc/polisade-orchestrator/init-verify.svg)](https://agentmods.dev/skills/cryndoc/polisade-orchestrator/init-verify)
Your own site
<a href="https://agentmods.dev/skills/cryndoc/polisade-orchestrator/init-verify"><img src="https://agentmods.dev/badge/skills/cryndoc/polisade-orchestrator/init-verify.svg" alt="Measured on agentmods" height="20"></a>
Per session 91 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,328 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5.1 $0.00091 $0.01328
Opus 5 $0.00046 $0.00664
Sonnet 5 $0.00018 $0.00266
Haiku 4.5 $0.00009 $0.00133

Measured 6d ago against content hash 26fb4fc5a398, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-06, from the pricing page.

Security

Grade A, and why

init-verify 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 6d 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.

skills/init-verify/SKILL.md · 122 lines

How it starts

The opening of the file, as written. The whole thing — 122 lines — stays where its author put it; the contents beside it link to each section on GitHub.

/polisade:init-verify — Verify Polisade Orchestrator init output

Guard-safe structural verification of a freshly initialized project. Reads only target-project files (never the plugin install directory), so it runs unblocked under the GigaCode Filesystem Guard. It exists to catch the issue #119 / #128 failure mode: a weak model that cannot Read the install dir silently reconstructs state files from memory (e.g. {"version": "5"} instead of {"polisadeVersion": "...", "schemaVersion": 7}, or an .env.example full of your_token_here). Those reconstructions are structurally wrong and this check makes them fail loud.

Algorithm

  1. Run the deterministic check below via run_shell_command. It reads only files in the current project (.state/*.json, the context file, .env.example) and prints exactly one PASS or FAIL: <reason> line.

    python3 - <<'PY'
    import json, os, sys
    
    # Bumped per release in lockstep with .claude-plugin/plugin.json. The
    # match is enforced by polisade_lint_skills.py::check_version_consistency
    # (invariant #1, 5th source) so this literal cannot silently drift.
    EXPECTED_POLISADE_VERSION = "3.7.5"
    
    fails = []
    
    # --- .state/PROJECT_STATE.json: the primary reconstruction tripwire ---
    ps_path = ".state/PROJECT_STATE.json"
    state = None
    if not os.path.isfile(ps_path):
        fails.append(f"{ps_path} is missing")
    else:
        try:
            state = json.load(open(ps_path, encoding="utf-8"))
        except (json.JSONDecodeError, OSError) as e:
            fails.append(f"{ps_path} is not valid JSON ({e})")
    if isinstance(state, dict):
        # Reconstructed files use a bare top-level `version` key — reject it.
        if "version" in state:
            fails.append(
                f"{ps_path} has a foreign top-level 'version' key "
                "(canonical schema uses polisadeVersion + schemaVersion)"
            )
        # polisadeVersion must EQUAL the current release — not just look like a
        # semver. A stale `2.24.1` in a `2.24.2` install means the file was
        # reconstructed (or a version-lockstep break) and must fail loud.
        pv = state.get("polisadeVersion")
        if pv != EXPECTED_POLISADE_VERSION:
            fails.append(
                f"{ps_path} polisadeVersion {pv!r} != expected "
                f"{EXPECTED_POLISADE_VERSION!r} (stale or reconstructed)"
            )
        if state.get("schemaVersion") != 7:
            fails.append(
                f"{ps_path} schemaVersion != 7 (got {state.get('schemaVersion')!r})"
            )
    
    # --- other .state JSON files must parse and exist ---
    for rel in (".state/counters.json", ".state/knowledge.json"):
        if not os.path.isfile(rel):
            fails.append(f"{rel} is missing")
            continue
        try:
            json.load(open(rel, encoding="utf-8"))
        except (json.JSONDecodeError, OSError) as e:
            fails.append(f"{rel} is not valid JSON ({e})")
    
    # --- context file present and non-trivial ---
    ctx = next((f for f in ("GIGACODE.md", "QWEN.md", "CLAUDE.md")
                if os.path.isfile(f)), None)
    if ctx is None:
        fails.append("context file (CLAUDE.md / QWEN.md / GIGACODE.md) is missing")
    elif os.path.getsize(ctx) < 200:
        fails.append(f"{ctx} is suspiciously small ({os.path.getsize(ctx)} bytes)")
    
    # --- provider-conditional .env.example check (explicit, never skipped) ---
    provider = None
    if isinstance(state, dict):
        provider = (state.get("settings") or {}).get("vcsProvider")
    if provider is None:
        fails.append(
            "settings.vcsProvider absent from PROJECT_STATE.json — cannot "
            "determine provider (likely reconstruction)"
        )
    elif provider == "bitbucket-server":
        ee = ".env.example"
        if not os.path.isfile(ee):
            fails.append(f"{ee} is missing for bitbucket-server provider")
        else:
            txt = open(ee, encoding="utf-8").read()
            for needle in ("BITBUCKET_DOMAIN1_URL", "BITBUCKET_DOMAIN2_URL"):
                if needle not in txt:
                    fails.append(f"{ee} missing canonical key {needle}")
            if "your_token_here" in txt:
                fails.append(f"{ee} contains placeholder 'your_token_here' (reconstructed)")
    
    if fails:
        for f in fails:
            print(f"FAIL: {f}")
        print(
            "FAIL: project structure looks RECONSTRUCTED, not written from "
            "canonical bytes. Re-run /polisade:init — do NOT invent or paraphrase "
            "the files. STOP and report; do not print the INITIALIZED banner."
        )
        sys.exit(1)
    print("PASS: project structure is canonical")
    PY
    

Read the full file on GitHub · 122 lines

Changes

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.

  1. 6d ago First seen · 122 lines · 91 tokens per session scan A 26fb4fc5a398

Subscribe to this mod's changes

init-verify is a skill published in the GitHub repository cryndoc/polisade-orchestrator (7 stars, last pushed 11d ago), licensed Apache-2.0. It adds 91 tokens to every session and 1,328 once invoked, about $0.0005 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 2026-08-31.

Related

Other skills, from other repositories

flow-lean

Lean output mode that fuses three disciplines: minimal solution (ponytail), action-first structure (adhd), zero-fat density (caveman). One rule under all of it: every token earns its place. A token earns its place only if it carries an action, a step, a decision-changing risk, or a proof. Everything else is cut.…

FlorianBruniaux/flow-lean · 89 tokens

deep-research-tiered

Deep research harness: fan-out search, fetch, adversarially verify, synthesize a cited report. Prefer Workflow({name:'deep-research-tiered', args:{question, models?}}) over built-in deep-research — same output, cheap tiered defaults (scope/search=sonnet, fetch=haiku, verify=sonnet, synthesize=inherit). Use for deep…

anshss/shiploop · 0 tokens

controllers-policy

Autonomy contract and routing index for autonomous project controllers. Loaded every tick; carries the rules that must survive cron-prompt rewrites.

Th0rgal/sandboxed.sh · 30 tokens

hermes-mission-control

How Hermes monitors and steers long-running sandboxed.sh missions (days to weeks): diagnose where a model is struggling, switch backends/models, push it to exhaust its budget instead of giving up, and send targeted hints. Trigger terms: mission, sandboxed.sh, babysit, monitor, /goal, switch backend, stalled, resume…

Th0rgal/sandboxed.sh · 94 tokens

open-dynamic-workflows

Plan, orchestrate, and adversarially verify parallel AI coding agents — a dynamic multi-agent workflow engine.

Suraj1235/open-dynamic-workflows · 28 tokens

ultracode

Ultracode-style dynamic workflows for Antigravity. Use when the user says "ultracode", "workflow:", "/deep-research", or asks for broad multi-file work with planning, parallel agents, verification, or crash-resume.

Suraj1235/open-dynamic-workflows · 56 tokens