wshobson--agents
be57c0b2e3
* feat(adapters): multi-harness framework + harness_portability eval dimension
Turn this Claude Code plugin marketplace into a generic agentic-harness
marketplace. Adapters under tools/adapters/ emit harness-native artifacts
for OpenAI Codex CLI, Cursor, OpenCode, and Gemini CLI from a single
Markdown source. Source-of-truth stays under plugins/ — Claude Code is
unchanged.
Framework (tools/adapters/):
- base.py — PluginSource parser, HarnessAdapter ABC, write/mirror helpers
(path-traversal guard, UTF-8-safe), inline-list + block-list + block-scalar
YAML-ish parser, _utf8_safe_cut, _split_inline_list, _normalize_author
- capabilities.py — per-harness capability matrix, TOOL_NAME_MAPS,
MODEL_ALIASES, resolve_model() with explicit warnings
- codex.py — emits .codex/{skills,agents}/ + AGENTS.md (≤150-line
table-of-contents). Fence-aware body splitter, _utf8_safe_cut for
multibyte safety, _yaml_scalar with reserved-word + special-char quoting.
Skill/command name collision detection (and second-order __cmd fallback).
- cursor.py — emits .cursor-plugin/{plugin,marketplace}.json + curated
.cursor/rules/*.mdc. _validate_mdc_frontmatter handles YAML block scalars
(no false positives on colons in description body). _normalize_author
handles dict, npm-style strings, and author lists.
- opencode.py — transpiles agents to .opencode/agents/<id>.md with
mode:subagent + permission: deny-everything-else block (skill/task always
allowed as base capabilities — Claude's implicit defaults).
- gemini.py — emits native skills/, agents/, and commands/ at extension
root (April 2026 spec). Tool-allowlist remapped via TOOL_NAME_MAPS.
CLI + tooling:
- tools/generate.py — unified `make generate HARNESS=<x> [PLUGIN=<y>]`,
with --clean (containment-guarded; case-insensitive on Darwin/Win32),
--prune (orphan removal across all per-harness output trees), --strict
(warnings fail), per-plugin error aggregation, refuses --clean --plugin
(would silently wipe other plugins' artifacts).
- tools/validate_generated.py — structural validation across all four
harness outputs. Codex 8KB cap → error. _extract_permission_block
correctly handles nested permission keys (column-0 only).
- tools/doc_gardener.py — recurring drift detection per OpenAI harness-
engineering principle. STALE_ARTIFACT (info), DEAD_LINK (error),
MARKETPLACE_ORPHAN (error), SKILL_OVER_CODEX_CAP (warning), grouped
output sorted by severity.
plugin-eval (extends existing framework):
- New harness_portability dimension (6% weight, rebalanced from existing
static sub-scores). Surfaces non-portable patterns with concrete
remediation hints: SKILL_OVER_CODEX_CAP, CLAUDE_TOOL_REFS,
CLAUDE_TOOL_PROSE, AGENT_NAME_COLLISION, BARE_MODEL_ALIAS.
- _CAMEL_TOOL_PATTERN requires Claude-tool context (no false positives
on Rust's `Task` etc.). _TOOL_PROSE_PATTERN case-sensitive on tool
names, case-insensitive on the leading article.
- Findings do NOT also feed anti_pattern_penalty (no double-counting).
Documentation:
- Top-level guides: CODEX.md, CURSOR.md, OPENCODE.md (≤150 lines each,
table-of-contents pattern per OpenAI harness-engineering post)
- docs/harnesses.md — capability matrix, graceful-degradation table,
generated output paths
- docs/authoring.md — portable-content style guide (tools, models,
collision rules, fence-respect)
- docs/round-trip-results.md — real-CLI verification recipes (OpenCode
discovers 193 subagents, Gemini extensions validate passes, Codex
TOMLs all parse)
- CONTRIBUTING.md — new file pointing at docs/authoring.md
- README.md — rewritten for multi-harness (145 lines, was 460)
- CLAUDE.md — trimmed to 60-line table-of-contents
- GEMINI.md — trimmed from 1500 to 500 tokens (3× over budget previously)
Tests: 181 passing (103 plugin-eval + 78 tools/tests). Real-CLI round-trip
verified for OpenCode, Gemini, and Codex (TOML parses).
Replaces tools/generate_gemini_commands.py with the unified CLI.
* refactor(skills): extract detail to references/details.md (~75 skills)
Apply Anthropic's canonical SKILL.md progressive-disclosure pattern across
the marketplace: SKILL.md body becomes a navigation tier (trigger phrasing
+ quick start), detailed templates and worked examples move to
references/details.md (loaded on demand by the agent).
Motivation: OpenAI Codex CLI hard-truncates skills at 8 KB. Before this
change, ~90 skills exceeded that cap and would silently break on Codex.
The progressive-disclosure pattern is also Anthropic's documented
recommendation for token efficiency — Claude Code reads references/ files
on demand when the body navigation says to.
What's extracted, by pattern:
- Pass 1 (## Templates section): 19 skills — full template libraries
moved to references/details.md
- Pass 2 (## Implementation Patterns / ## Advanced Patterns): 13 skills
- Pass 3 (everything between nav-tier and wrap-tier headings): 53 skills
- Conservative re-extraction for 8 skills that got over-reduced — kept
~6-7 KB inline (most of the quick-start tier) plus references/ overflow
What stays inline (SKILL.md navigation tier):
- description: frontmatter (triggering — unchanged for all skills)
- ## When to Use This Skill / ## Core Concepts / ## Quick Start
- ## Best Practices / ## Troubleshooting / ## See Also wrap-ups
- A pointer note ("see references/details.md") so the agent knows where
to look for detail
What goes to references/details.md (detail tier, on-demand load):
- ## Templates (full code template libraries)
- ## Implementation Patterns / ## Advanced Patterns (deep examples)
- Mid-skill walkthroughs that exceed the inline budget
Also in this commit:
- plugins/brand-landingpage description trimmed from 958→543 chars
(preserves trigger phrasing, drops verbose example-quote list)
Net effect:
- SKILL_OVER_CODEX_CAP findings: 90 → 10 (88% reduction)
- All triggers unchanged — discovery behavior identical across harnesses
- 75 new references/details.md files with the extracted content
- Same depth of guidance, loaded progressively
Remaining 10 oversized skills are complex multi-section docs (e.g.
postgresql, code-review-excellence, evaluation-methodology) that need
per-skill manual judgment — flagged by `make garden` for future work.
* chore: bump all plugin versions (multi-harness release)
Patch-bump every local plugin (81) in both .claude-plugin/marketplace.json
entries and each plugins/<name>/.claude-plugin/plugin.json. Minor-bump the
top-level marketplace metadata.version (1.6.0 → 1.7.0) to signal the
multi-harness adapter framework addition.
The external git-subdir entry (qa-orchestra) is unaffected — its version
is governed by its upstream repo.
* fix(opencode): preserve explicit tools:[] + word-boundary subtask match
Addresses two Codex review findings on PR #541.
## P1 — `tools: []` silently upgraded to permissive (privilege escalation)
Before: `_build_permission_block` returned `{}` for any empty list, which
omits the `permission:` block entirely from the emitted agent. An author
who explicitly wrote `tools: []` to lock down an advisory-only agent got
an UNRESTRICTED agent in OpenCode. Affected agent in this tree:
`plugins/arm-cortex-microcontrollers/agents/arm-cortex-expert.md`.
Fix: `_build_permission_block` now takes a `has_tools_field` flag so the
caller can distinguish "tools: key missing" (Claude default permissive)
from "tools: []" (explicit lock-down). The lock-down case emits a
deny-everything block that allows ONLY the base capabilities (skill, task)
that Claude Code always grants implicitly. Verified against the real
arm-cortex-expert agent — now emits read/edit/write/bash/grep/glob/list:
deny, task/skill: allow.
## P2 — `"agent" in cmd.body.lower()` false-positives on substrings
Before: a command body containing `PerformanceReviewAgent` (class name
in a code snippet) or `useragent` triggered `subtask: true`, changing
runtime behavior based on incidental text.
Fix: switch to a compiled word-boundary regex `\b(agent|subagent)s?\b`
(case-insensitive). Tests confirm the substring `PerformanceReviewAgent`
no longer fires, while a real "spawn a subagent" sentence still does.
## Tests
3 new regression tests in tools/tests/test_adapters.py:
- `test_explicit_empty_tools_yields_locked_permission_block` (P1)
- `test_missing_tools_field_yields_no_permission_block` (P1 boundary)
- `test_subtask_inference_word_boundary` (P2)
184 total tests pass (was 181). OpenCode round-trip still discovers all
193 subagents; arm-cortex-expert agent is now properly locked down.
* test: behavioral verification + CI gates for multi-harness pipeline
Adds three layers of automated verification that pure-Python parser tests
miss, plus the CI jobs that turn them into hard gates. Catches the kinds
of issues that previously only surfaced when a real user installed the
marketplace and tried to use it.
## test_real_world.py — real-source structural tests
Runs against the actual `plugins/` tree (not synthetic fixtures). Catches
issues that only appear on real content:
- every marketplace entry resolves to a plugins/<name>/ dir
- every local plugin dir appears in marketplace.json
- marketplace.json version == per-plugin plugin.json version (catches drift)
- every plugin loads via load_plugin() without error
- no plugin name contains `__` (adapter namespace separator)
- every agent has name + description; every skill has a trigger phrase
(same regex plugin_eval's MISSING_TRIGGER check uses)
- no agent name collides with Codex built-ins
- every refactored skill (with `references/details.md`) has:
- meaningful detail content (>=500 B in details.md)
- a pointer to references/ in the SKILL.md body
- a navigation-tier heading preserved (When to Use, Overview, etc.)
- body >= 600 B (not a stub)
- every plugin.json has name + version matching the dir
This test pass found and fixed three real defects before commit:
- ship-mate/skills/scan: description had no trigger phrase ("Use when…")
- reverse-engineering/skills/memory-forensics: nav-tier section lost
during extraction
- reverse-engineering/skills/binary-analysis-patterns: same
All three are now fixed (preserved trigger phrasing, added When-to-Use
sections back to the skills my extraction over-trimmed).
## test_round_trip.py — generate→parse→verify
CI runs this AFTER `make generate-all`. Catches generation-time regressions:
- OpenCode/Codex/Gemini agent counts match source agent count (no skips)
- every Codex SKILL.md under 8 KB (the cap that would silently truncate)
- every Codex agent TOML has required fields + valid sandbox_mode
- every OpenCode agent has mode in {primary,subagent,all} and
provider-prefixed model
- locked agents (source `tools: []`) emit proper deny-everything permission
block with skill/task allow (regression guard for PR-541 P1)
- every Gemini @{path} injection resolves to a real source file
- every Gemini command TOML has prompt + {{args}} placeholder
- every context file (CLAUDE.md, AGENTS.md, GEMINI.md, etc.) within
150-line cap
- Cursor marketplace + per-plugin manifests cover all local plugins
- .cursor/rules/*.mdc only use the 3 documented frontmatter keys
## test_cli_smoke.py — real-CLI subprocess tests
Invokes the actual harness binaries (OpenCode, Gemini, Codex, Claude Code)
against the generated artifacts. Catches CLI-level issues pure-Python
parsing can't see: schema-loader drift, plugin-discovery bugs, version
incompatibilities.
- `opencode agent list` — must succeed AND discover every source agent
(currently 191 + 2 OpenCode built-ins)
- `gemini extensions validate <repo>` — must return success
- `codex doctor` — must report healthy install
- every Codex agent TOML must parse with stdlib `tomllib`
- `claude --version` — sanity check the Claude Code CLI loads
- marketplace.json must have owner + metadata.version for Claude Code's loader
Per-CLI tests skip gracefully when the binary isn't on PATH, so local
devs only exercise what they have installed. CI installs OpenCode +
Gemini and turns those skips into hard gates.
## Makefile + CI
- `make test` — full pytest suite (plugin-eval + tools/tests/)
- `make smoke-test` — generates if needed, then runs real-CLI smoke tests
- `.github/workflows/validate.yml` extended with:
- `tools-tests` job — runs pytest tools/tests/
- `multi-harness-generate` job — `make generate-all && make validate
STRICT=1 && make garden`, uploads generated artifacts on every run
- `cli-smoke-test` job — installs OpenCode + Gemini, runs test_cli_smoke.py
## Test counts
- Before: 184 tests
- After: 386 tests (parameterized real-source tests over all 82 plugins)
- All passing locally on OpenCode 1.15.7 + Gemini 0.42.0 + Codex 0.133.0
+ Claude Code 2.1.148
12 KiB
12 KiB
error-handling-patterns — detailed patterns and worked examples
Language-Specific Patterns
Python Error Handling
Custom Exception Hierarchy:
class ApplicationError(Exception):
"""Base exception for all application errors."""
def __init__(self, message: str, code: str = None, details: dict = None):
super().__init__(message)
self.code = code
self.details = details or {}
self.timestamp = datetime.utcnow()
class ValidationError(ApplicationError):
"""Raised when validation fails."""
pass
class NotFoundError(ApplicationError):
"""Raised when resource not found."""
pass
class ExternalServiceError(ApplicationError):
"""Raised when external service fails."""
def __init__(self, message: str, service: str, **kwargs):
super().__init__(message, **kwargs)
self.service = service
# Usage
def get_user(user_id: str) -> User:
user = db.query(User).filter_by(id=user_id).first()
if not user:
raise NotFoundError(
f"User not found",
code="USER_NOT_FOUND",
details={"user_id": user_id}
)
return user
Context Managers for Cleanup:
from contextlib import contextmanager
@contextmanager
def database_transaction(session):
"""Ensure transaction is committed or rolled back."""
try:
yield session
session.commit()
except Exception as e:
session.rollback()
raise
finally:
session.close()
# Usage
with database_transaction(db.session) as session:
user = User(name="Alice")
session.add(user)
# Automatic commit or rollback
Retry with Exponential Backoff:
import time
from functools import wraps
from typing import TypeVar, Callable
T = TypeVar('T')
def retry(
max_attempts: int = 3,
backoff_factor: float = 2.0,
exceptions: tuple = (Exception,)
):
"""Retry decorator with exponential backoff."""
def decorator(func: Callable[..., T]) -> Callable[..., T]:
@wraps(func)
def wrapper(*args, **kwargs) -> T:
last_exception = None
for attempt in range(max_attempts):
try:
return func(*args, **kwargs)
except exceptions as e:
last_exception = e
if attempt < max_attempts - 1:
sleep_time = backoff_factor ** attempt
time.sleep(sleep_time)
continue
raise
raise last_exception
return wrapper
return decorator
# Usage
@retry(max_attempts=3, exceptions=(NetworkError,))
def fetch_data(url: str) -> dict:
response = requests.get(url, timeout=5)
response.raise_for_status()
return response.json()
TypeScript/JavaScript Error Handling
Custom Error Classes:
// Custom error classes
class ApplicationError extends Error {
constructor(
message: string,
public code: string,
public statusCode: number = 500,
public details?: Record<string, any>,
) {
super(message);
this.name = this.constructor.name;
Error.captureStackTrace(this, this.constructor);
}
}
class ValidationError extends ApplicationError {
constructor(message: string, details?: Record<string, any>) {
super(message, "VALIDATION_ERROR", 400, details);
}
}
class NotFoundError extends ApplicationError {
constructor(resource: string, id: string) {
super(`${resource} not found`, "NOT_FOUND", 404, { resource, id });
}
}
// Usage
function getUser(id: string): User {
const user = users.find((u) => u.id === id);
if (!user) {
throw new NotFoundError("User", id);
}
return user;
}
Result Type Pattern:
// Result type for explicit error handling
type Result<T, E = Error> = { ok: true; value: T } | { ok: false; error: E };
// Helper functions
function Ok<T>(value: T): Result<T, never> {
return { ok: true, value };
}
function Err<E>(error: E): Result<never, E> {
return { ok: false, error };
}
// Usage
function parseJSON<T>(json: string): Result<T, SyntaxError> {
try {
const value = JSON.parse(json) as T;
return Ok(value);
} catch (error) {
return Err(error as SyntaxError);
}
}
// Consuming Result
const result = parseJSON<User>(userJson);
if (result.ok) {
console.log(result.value.name);
} else {
console.error("Parse failed:", result.error.message);
}
// Chaining Results
function chain<T, U, E>(
result: Result<T, E>,
fn: (value: T) => Result<U, E>,
): Result<U, E> {
return result.ok ? fn(result.value) : result;
}
Async Error Handling:
// Async/await with proper error handling
async function fetchUserOrders(userId: string): Promise<Order[]> {
try {
const user = await getUser(userId);
const orders = await getOrders(user.id);
return orders;
} catch (error) {
if (error instanceof NotFoundError) {
return []; // Return empty array for not found
}
if (error instanceof NetworkError) {
// Retry logic
return retryFetchOrders(userId);
}
// Re-throw unexpected errors
throw error;
}
}
// Promise error handling
function fetchData(url: string): Promise<Data> {
return fetch(url)
.then((response) => {
if (!response.ok) {
throw new NetworkError(`HTTP ${response.status}`);
}
return response.json();
})
.catch((error) => {
console.error("Fetch failed:", error);
throw error;
});
}
Rust Error Handling
Result and Option Types:
use std::fs::File;
use std::io::{self, Read};
// Result type for operations that can fail
fn read_file(path: &str) -> Result<String, io::Error> {
let mut file = File::open(path)?; // ? operator propagates errors
let mut contents = String::new();
file.read_to_string(&mut contents)?;
Ok(contents)
}
// Custom error types
#[derive(Debug)]
enum AppError {
Io(io::Error),
Parse(std::num::ParseIntError),
NotFound(String),
Validation(String),
}
impl From<io::Error> for AppError {
fn from(error: io::Error) -> Self {
AppError::Io(error)
}
}
// Using custom error type
fn read_number_from_file(path: &str) -> Result<i32, AppError> {
let contents = read_file(path)?; // Auto-converts io::Error
let number = contents.trim().parse()
.map_err(AppError::Parse)?; // Explicitly convert ParseIntError
Ok(number)
}
// Option for nullable values
fn find_user(id: &str) -> Option<User> {
users.iter().find(|u| u.id == id).cloned()
}
// Combining Option and Result
fn get_user_age(id: &str) -> Result<u32, AppError> {
find_user(id)
.ok_or_else(|| AppError::NotFound(id.to_string()))
.map(|user| user.age)
}
Go Error Handling
Explicit Error Returns:
// Basic error handling
func getUser(id string) (*User, error) {
user, err := db.QueryUser(id)
if err != nil {
return nil, fmt.Errorf("failed to query user: %w", err)
}
if user == nil {
return nil, errors.New("user not found")
}
return user, nil
}
// Custom error types
type ValidationError struct {
Field string
Message string
}
func (e *ValidationError) Error() string {
return fmt.Sprintf("validation failed for %s: %s", e.Field, e.Message)
}
// Sentinel errors for comparison
var (
ErrNotFound = errors.New("not found")
ErrUnauthorized = errors.New("unauthorized")
ErrInvalidInput = errors.New("invalid input")
)
// Error checking
user, err := getUser("123")
if err != nil {
if errors.Is(err, ErrNotFound) {
// Handle not found
} else {
// Handle other errors
}
}
// Error wrapping and unwrapping
func processUser(id string) error {
user, err := getUser(id)
if err != nil {
return fmt.Errorf("process user failed: %w", err)
}
// Process user
return nil
}
// Unwrap errors
err := processUser("123")
if err != nil {
var valErr *ValidationError
if errors.As(err, &valErr) {
fmt.Printf("Validation error: %s\n", valErr.Field)
}
}
Universal Patterns
Pattern 1: Circuit Breaker
Prevent cascading failures in distributed systems.
from enum import Enum
from datetime import datetime, timedelta
from typing import Callable, TypeVar
T = TypeVar('T')
class CircuitState(Enum):
CLOSED = "closed" # Normal operation
OPEN = "open" # Failing, reject requests
HALF_OPEN = "half_open" # Testing if recovered
class CircuitBreaker:
def __init__(
self,
failure_threshold: int = 5,
timeout: timedelta = timedelta(seconds=60),
success_threshold: int = 2
):
self.failure_threshold = failure_threshold
self.timeout = timeout
self.success_threshold = success_threshold
self.failure_count = 0
self.success_count = 0
self.state = CircuitState.CLOSED
self.last_failure_time = None
def call(self, func: Callable[[], T]) -> T:
if self.state == CircuitState.OPEN:
if datetime.now() - self.last_failure_time > self.timeout:
self.state = CircuitState.HALF_OPEN
self.success_count = 0
else:
raise Exception("Circuit breaker is OPEN")
try:
result = func()
self.on_success()
return result
except Exception as e:
self.on_failure()
raise
def on_success(self):
self.failure_count = 0
if self.state == CircuitState.HALF_OPEN:
self.success_count += 1
if self.success_count >= self.success_threshold:
self.state = CircuitState.CLOSED
self.success_count = 0
def on_failure(self):
self.failure_count += 1
self.last_failure_time = datetime.now()
if self.failure_count >= self.failure_threshold:
self.state = CircuitState.OPEN
# Usage
circuit_breaker = CircuitBreaker()
def fetch_data():
return circuit_breaker.call(lambda: external_api.get_data())
Pattern 2: Error Aggregation
Collect multiple errors instead of failing on first error.
class ErrorCollector {
private errors: Error[] = [];
add(error: Error): void {
this.errors.push(error);
}
hasErrors(): boolean {
return this.errors.length > 0;
}
getErrors(): Error[] {
return [...this.errors];
}
throw(): never {
if (this.errors.length === 1) {
throw this.errors[0];
}
throw new AggregateError(
this.errors,
`${this.errors.length} errors occurred`,
);
}
}
// Usage: Validate multiple fields
function validateUser(data: any): User {
const errors = new ErrorCollector();
if (!data.email) {
errors.add(new ValidationError("Email is required"));
} else if (!isValidEmail(data.email)) {
errors.add(new ValidationError("Email is invalid"));
}
if (!data.name || data.name.length < 2) {
errors.add(new ValidationError("Name must be at least 2 characters"));
}
if (!data.age || data.age < 18) {
errors.add(new ValidationError("Age must be 18 or older"));
}
if (errors.hasErrors()) {
errors.throw();
}
return data as User;
}
Pattern 3: Graceful Degradation
Provide fallback functionality when errors occur.
from typing import Optional, Callable, TypeVar
T = TypeVar('T')
def with_fallback(
primary: Callable[[], T],
fallback: Callable[[], T],
log_error: bool = True
) -> T:
"""Try primary function, fall back to fallback on error."""
try:
return primary()
except Exception as e:
if log_error:
logger.error(f"Primary function failed: {e}")
return fallback()
# Usage
def get_user_profile(user_id: str) -> UserProfile:
return with_fallback(
primary=lambda: fetch_from_cache(user_id),
fallback=lambda: fetch_from_database(user_id)
)
# Multiple fallbacks
def get_exchange_rate(currency: str) -> float:
return (
try_function(lambda: api_provider_1.get_rate(currency))
or try_function(lambda: api_provider_2.get_rate(currency))
or try_function(lambda: cache.get_rate(currency))
or DEFAULT_RATE
)
def try_function(func: Callable[[], Optional[T]]) -> Optional[T]:
try:
return func()
except Exception:
return None