Helpers
Persistent helper agents for research, audit, quality, and orchestration with frontmatter contracts and deployment isolation protocols.
Persistent helpers live as flat files under the canonical helpers
directory with the <name>.md suffix. The canonical path is:
src/apothem/agents/Each helper is a focused role, such as research, audit, quality, refactoring, or scaffolding, with its own tool restrictions and system prompt.
The table and the path list below show four of the helpers. The generated inventory at the end of this page lists every helper definition.
| Helper | File | Pattern | Model Policy |
|---|---|---|---|
| Codebase Explorer | see manifest path below | Research | host-selected |
| Convention Auditor | see manifest path below | Audit | host-selected |
| Quality Gate | see manifest path below | Quality | host-selected |
| Memory Auditor | see manifest path below | Audit | host-selected |
Canonical manifest paths:
src/apothem/agents/codebase-explorer.md
src/apothem/agents/convention-auditor.md
src/apothem/agents/quality-gate.md
src/apothem/agents/memory-auditor.mdFrontmatter contract
---
name: <agent name>
version: <semver>
updated: <YYYY-MM-DD>
description: <one-line role>
tools: <comma-separated tool list>
disallowedTools: <comma-separated deny list>
maxTurns: <integer>
---tools is optional. A helper that omits it inherits every tool the session has,
connected MCP tools included, and disallowedTools carries its limit. All
fourteen shipped helpers list tools.
Connected MCP servers
A helper that lists tools can call only what it lists. Claude Code admits an
MCP tool to that list only when the entry names the server, as in
mcp__<server>__*; a bare mcp__* is documented as a disallowedTools entry
and nowhere else. Server names belong to the person who connects them, so a
shipped helper cannot name them.
citation-auditor and research-scout look up scholarly sources. Each ships
tools: Read, Glob, Grep, WebSearch, WebFetch and denies the tools that write
files, run commands, start other helpers, or publish: Write, Edit,
MultiEdit, NotebookEdit, Bash, PowerShell, Monitor, Agent,
Artifact, and the task-list tools. With no profile setting, neither reaches an
MCP tool, and both look sources up through web search and web fetch.
To give one of them a connected scholarly index, name the server in the
profile's agent_mcp_servers key, which maps an agent name to server names. The
Claude Code engine install (apothem install) then appends one
mcp__<server>__* entry per named server to that agent's tools line, and
changes no other agent. The Gemini CLI install does the same with one
mcp_<server>_* entry per server in the agent's tools allowlist, for a server
name of letters, digits and -. On both, a server the agent's own
disallowedTools names stays denied. The entry admits every tool of that server, including
any that writes, so name only a server you trust for lookups: the helpers'
instructions limit them to search, look-up, read, verify, and resolve calls, but
the harness enforces only the names the tools line lists. A name for an agent
that does not exist, or that has no explicit tools line, is reported and
skipped.
Harnesses other than Claude Code and Gemini CLI ignore the key. The Claude Code plugin tree is assembled without
a profile and keeps the plain list, so a plugin user defines a user-level agent
under ~/.claude/agents/ instead. The profile syntax, the server-name rules, and
that workaround are on
Give a helper access to MCP servers.
Per-harness conversion
A helper whose tools grant, after disallowedTools, includes no file-editing
tool (Write, Edit, MultiEdit, or NotebookEdit) is read-only. A helper
with no tools grant is read-only when disallowedTools denies at least
Write and Edit. Each harness that converts helpers enforces that
limit natively, states it in the helper's instructions, or drops the read-only label:
| Harness | Tool limit | Turn limit | Model and effort |
|---|---|---|---|
| Claude Code | tools / disallowedTools as written, plus any mcp__<server>__* entries the profile's agent_mcp_servers key adds to a listed tools line; with no tools, every tool is inherited, MCP included | maxTurns | inherited (not set) |
| Codex | none emitted; the helper's instructions state it, and the parent session's sandbox governs writes | not available | inherited (not set) |
| Gemini CLI | tools allowlist of Gemini tool names, plus any mcp_<server>_* entries the profile's agent_mcp_servers key adds to a listed explicit list; with no tools, the allowlist the denials imply plus mcp_* | max_turns | inherited (not set) |
| Qwen Code | tools / disallowedTools in Qwen tool names; with no tools, no list is emitted, so MCP tools are inherited and the denials apply | not mapped | model: inherit |
| OpenCode | permission (allow / deny per tool); MCP tools are not restricted | not mapped | inherited (not set) |
| Antigravity | none emitted; the description drops "read-only" | not available | inherited (not set) |
| CodeBuddy | tools / disallowedTools as comma strings, limited to the tool names CodeBuddy documents; no MCP entry | maxTurns | inherited (not set) |
| Trae | tools / disallowedTools as comma strings, limited to the tool names Trae documents; no MCP entry | not available | inherited (not set) |
| GitHub Copilot | tools allowlist of Copilot aliases (read, search, edit, execute, web, agent, todo) with the denials folded in, because Copilot has no deny key; an alias that covers a denied tool is left out; no MCP entry | not available | inherited (not set) |
| Cursor | no tool list (Cursor has none); readonly: true for the five read-only helpers without Bash, and the other read-only helpers drop the "read-only" label | not available | inherited (not set) |
| Kimi Code | tools / disallowedTools as lists of Kimi Code tool names (WebFetch becomes FetchURL, TodoWrite becomes TodoList; names Kimi Code has no tool for are left out); no MCP entry | not available | inherited (not set) |
| Kiro | tools allowlist of Kiro tool categories (read, write, shell, web, subagent) with the denials folded in, because the list is the only restriction written; a category that covers a denied tool is left out; no MCP entry | not available | inherited (not set) |
| Windsurf | allowed-tools allowlist of Devin tool names (read, edit, grep, glob, exec, write) with the denials folded in, because Devin has no deny key; the three helpers that need web tools are not installed; no MCP entry | not available | Devin's default subagent model (not set) |
Hermes and Open-Claw receive the helper files as support material, not as native helpers.
Orchestration
Helper deployment, parallel-launch protocol, return-contract enforcement, and isolation discipline live at the orchestration rule. The per-worker dispatch tree lives at the orchestration patterns rule, §Decision Tree. Canonical paths:
src/apothem/rules/agent-orchestration.md
src/apothem/rules/agent-orchestration-patterns.mdGenerated Agent Inventory
| Name | Description | Source |
|---|---|---|
README | Persistent agent definitions — the reusable sub-agent specifications the harness materializes whenever the orchestrator deploys a research, audit, or quality team. Each agent is a | src/apothem/agents/README.md |
citation-auditor | Read-only citation audit — checks that each reference in a bibliography or cited passage exists, matches its record, and that the opened source states what the text attributes to it. Use when a draft cites sources: 'check my references', 'are these DOIs real', 'does this paper say what I cite it for'. Finding new sources routes to research-scout; a free-standing claim routes to fact-checker. | src/apothem/agents/citation-auditor.md |
codebase-explorer | Read-only codebase exploration — exhaustively find every instance of a pattern, trace call/import dependencies, discover host conventions, map architecture and layering. Use when the query is 'where is X used', 'find all callers of Y', 'what convention does this repo follow for Z', 'map the module structure', or 'trace what depends on this file' — never for edits or design, and never for review verdicts (architecture-review, code-audit). Works through Grep / Glob / Read plus read-only Bash and returns evidence-cited findings (path + line + snippet) exhaustive within the declared scope; reports gaps rather than guessing. | src/apothem/agents/codebase-explorer.md |
convention-auditor | Read-only conventions auditor: verify naming (kebab-case files, snake_case Python, frontmatter keys), cross-reference resolution (dead rule/skill/agent paths, broken anchors, half-edge bindings), formatting standards (SPDX header presence, frontmatter contract), and ecosystem coherence (orphan artifacts, sibling-convention drift, registry gaps). Dispatch when a change touches multiple rules/skills/agents/commands and you need a binary PASS/FINDING verdict per instance with file+line+expected+actual evidence — e.g. 'audit naming across the harness adapters', 'check every new rule resolves its Bindings cross-refs', 'verify the skills cohort matches the kebab-case + SKILL.md convention'. Read-only: never fixes, never runs shell. A whole-tree sweep, or any fix, is the ecosystem-audit skill. Conventions are owned upstream and verified against, never re-derived; every finding cites severity + evidence. | src/apothem/agents/convention-auditor.md |
dependency-auditor | Read-only dependency-tree risk audit — flags unpinned, stale, duplicate, and known-vulnerable dependencies with manifest/lockfile evidence. Dispatch within a dependency audit, before a release cut, or after adding a dependency; one named CVE or advisory routes to the vuln-triage skill, and a signing or provenance review to the supply-chain-audit command. Detects the ecosystem via host-discovery: npm (package.json + package-lock/yarn/pnpm), pip (pyproject/requirements + lock), cargo (Cargo.toml + Cargo.lock), go (go.mod + go.sum). Runs npm audit / pip-audit / cargo audit / govulncheck and returns a risk-ranked table (known-vulnerable → unpinned → duplicate → stale) with per-row evidence; never widens a pin or escalates a scope — gaps surface as inquiry. | src/apothem/agents/dependency-auditor.md |
design-critic | Read-only UI critique. Use for a design review, a usability or accessibility check, or 'what is wrong with this screen' before a UI ships. A WCAG sweep across a whole web surface routes to the a11y-audit command; a CLI or install-flow review to ux-review. Reviews code, screenshots, or a running page and returns fixes ranked by user impact, from a scored heuristic review and separately measured evidence. Opens with a degraded notice when measurement tooling is missing; never guesses a measurement. | src/apothem/agents/design-critic.md |
fact-checker | Read-only adversarial claim verification — decompose input into atomic claims, seek ≥2 independent sources, attempt refutation, assign cited verdicts (supported / refuted / unverifiable) with quoted evidence and confidence. Use when a claim needs proof before it ships: a benchmark or statistic in docs/copy, a 'X is faster/safer than Y' assertion, a citation that names an RFC or spec, a release note, or any factual claim a reviewer would challenge. Routes external claims through WebSearch / WebFetch and repository claims through Read / Glob / Grep; defaults to refuted-or-unverifiable when evidence is insufficient, never a charitable supported. | src/apothem/agents/fact-checker.md |
mcp-builder | Scaffold a Model Context Protocol (MCP) server skeleton from a tool/resource contract — contract-first, well-typed tools, minimal surface. Use when: 'build an MCP server for <API>', 'scaffold MCP tools from this spec', 'wire FastMCP/TypeScript-SDK tool definitions', 'add a tool that exposes <resource> over MCP'. Detection: tool names + argument shapes + return types + resource URIs are stated or derivable. Selects the SDK via host-discovery (FastMCP for Python, the TypeScript SDK for Node), emits one typed tool definition per contract entry plus a list-tools smoke test, and scaffolds nothing speculative. Not for: tuning, securing, or load-testing an existing server (those surface as adjacent gaps). | src/apothem/agents/mcp-builder.md |
memory-auditor | Read-only memory-file auditor: cross-reference every claim in the harness memory tier (MEMORY.md index + topic files under <harness-root>/projects/{hash}/memory/ and <harness-root>/memory/) against actual filesystem state — file/line/rule counts (glob and count), referenced paths (do they exist?), rule scope labels (match pathFilter frontmatter?), dates (against frontmatter updated: or mtime, never the system clock), and cross-reference matrices. Dispatch when MEMORY.md or a topic file may have drifted from reality and you need a per-claim PASS/FAIL verdict with contradicting evidence — e.g. 'audit MEMORY.md after the rules cohort was renamed', 'verify the memory index counts match the current ecosystem', 'check the debugging topic file for stale references to deleted artifacts'. Existence + name match only; never re-audit an artifact's internal correctness (convention-auditor's scope). Read-only: never writes, never fixes. | src/apothem/agents/memory-auditor.md |
prompt-evaluator | Read-only rubric scoring of a prompt's output set — score each output against each named criterion (PASS/FAIL with cited evidence), aggregate per-criterion pass-rate, flag regressions against a baseline, and name recurring failure modes. Use when prompt or model outputs need a reproducible verdict before they ship: comparing a reprompted variant against the prior version, gating a system-prompt change, judging an LLM-as-judge eval set, scoring few-shot outputs against acceptance criteria, or proving a quality claim a reviewer would challenge. Scores outputs that already exist: building the dataset, scorer or runner routes to the eval-harness skill, and a full campaign to the eval command. Requires an explicit rubric — an absent or underspecified rubric blocks scoring and routes as inquiry; never invents a criterion or threshold, never scores charitably. | src/apothem/agents/prompt-evaluator.md |
quality-gate | Read-only quality-gate runner — discovers the host's lint / type-check / test / security / build commands, runs them in the correct order (build → type-check → tests+lint+security in parallel), and returns a per-gate PASS/FAIL verdict with file+line+error evidence. Reports, never fixes. Dispatch as a Quality team before a release cut, after a multi-file change, or to confirm a fix is green — e.g. 'run the full quality matrix and tell me what fails', 'gate this branch before I push', 'is the test suite green and the types clean?'. Runs the whole matrix; tests alone route to test-runner. Detects tooling via host-discovery (ruff/eslint/markdownlint, mypy/tsc/pyright, pytest/jest/cargo test/go test, bandit/npm audit/gitleaks); never assumes a stack. | src/apothem/agents/quality-gate.md |
refactor-surgeon | Scoped, behavior-preserving refactor of a named target — extract the behavioral contract, re-derive clean-room (never edit in place), name the one deficiency removed, verify regression via the host's own tests. Dispatch as a delegated team member on a single named target with a clear refactor intent; a user-driven interactive extraction is the refactor-extract skill — e.g. 'consolidate the duplicated validation in the CLI's install module', 'untangle the nested conditionals in the config renderer without changing output', 'rename the god-object methods in the adapter class to reveal intent'. Touches only the named target; adjacent gaps surface as findings, never as edits. Behavior, contracts, and side effects are identical before and after; a behavior change is a defect. | src/apothem/agents/refactor-surgeon.md |
research-scout | Read-only source discovery and ranking — decompose a research question into facets, fan out parallel web queries, rank candidates by authority, recency, and relevance, and return a deduplicated ranked source list. Use when a question needs sources before it can be answered: 'find the authoritative spec/RFC for X', 'what are the primary sources on Y', 'gather current references for a docs/copy claim', 'survey the landscape before a deep dive'. Fans external facets through any scholarly index server the operator has granted it, WebSearch, and WebFetch, and any in-repo corpus through Read / Glob / Grep. Also looks up papers and preprints by DOI, arXiv id, or topic ('find the paper behind this DOI'), and scouts existing open-source projects before anything new is built ('is there already a library for this', 'adopt or fork'). Discovery and ranking only — never fabricates a URL, never synthesizes; a written cited report routes to the multi-source-research skill, claim verification to fact-checker, and combining given documents to source-synthesis. | src/apothem/agents/research-scout.md |
security-scanner | Read-only secret, SAST-pattern, and config-risk sweep — a coarse first-pass that surfaces candidates fast and routes deep scanning to the host's CI scanners. Dispatch inside a security audit, after touching auth/crypto/IO surfaces, or when a secret may have been committed; a full repository security audit routes to the security-audit command. Greps for credential literals (key/token/password/private-key/certificate, committed .env values), injection surfaces (shell=True on untrusted input, string-formatted SQL, command interpolation), unsafe-eval (eval/exec/Invoke-Expression), unsafe deserialization (pickle.loads, yaml.load without SafeLoader), and config risk (over-broad CI permissions:, unpinned Actions uses:, network egress). Routes known-vulnerable dependencies to the dependency-auditor agent, CVE/advisory triage to the vuln-triage skill, and leaked-credential rotation to the secret-rotation skill. | src/apothem/agents/security-scanner.md |
test-runner | Read-only test-suite runner — discovers the host's test command, runs it, and triages every failure by root cause with test+file+line+assertion evidence. Reports, never fixes. Dispatch as a Research/Quality team member after a code change, before a release cut, or to confirm a fix is green — e.g. 'run the tests and tell me what's failing and why', 'is the suite green on this branch?', 'triage the failures in the auth module'. Tests only: lint, type and security gates route to quality-gate. Detects the runner via host-discovery (pytest/jest/cargo test/go test/Makefile target from the manifest + CI config); never assumes pytest. | src/apothem/agents/test-runner.md |
Commands
Slash-command definitions implementing the plan-suite pipeline plus audits for code, architecture, security, dependencies, and documentation review.
Hooks
Tool event handlers that fire on SessionStart, PreToolUse, PostToolUse, and Stop, loading message contexts enforcing git isolation and plan discipline, with a tested runtime contract and per-user state.