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
9.5 KiB
9.5 KiB
bash-defensive-patterns — detailed patterns and worked examples
Core Defensive Principles
1. Strict Mode
Enable bash strict mode at the start of every script to catch errors early.
#!/bin/bash
set -Eeuo pipefail # Exit on error, unset variables, pipe failures
Key flags:
set -E: Inherit ERR trap in functionsset -e: Exit on any error (command returns non-zero)set -u: Exit on undefined variable referenceset -o pipefail: Pipe fails if any command fails (not just last)
2. Error Trapping and Cleanup
Implement proper cleanup on script exit or error.
#!/bin/bash
set -Eeuo pipefail
trap 'echo "Error on line $LINENO"' ERR
trap 'echo "Cleaning up..."; rm -rf "$TMPDIR"' EXIT
TMPDIR=$(mktemp -d)
# Script code here
3. Variable Safety
Always quote variables to prevent word splitting and globbing issues.
# Wrong - unsafe
cp $source $dest
# Correct - safe
cp "$source" "$dest"
# Required variables - fail with message if unset
: "${REQUIRED_VAR:?REQUIRED_VAR is not set}"
4. Array Handling
Use arrays safely for complex data handling.
# Safe array iteration
declare -a items=("item 1" "item 2" "item 3")
for item in "${items[@]}"; do
echo "Processing: $item"
done
# Reading output into array safely
mapfile -t lines < <(some_command)
readarray -t numbers < <(seq 1 10)
5. Conditional Safety
Use [[ ]] for Bash-specific features, [ ] for POSIX.
# Bash - safer
if [[ -f "$file" && -r "$file" ]]; then
content=$(<"$file")
fi
# POSIX - portable
if [ -f "$file" ] && [ -r "$file" ]; then
content=$(cat "$file")
fi
# Test for existence before operations
if [[ -z "${VAR:-}" ]]; then
echo "VAR is not set or is empty"
fi
Fundamental Patterns
Pattern 1: Safe Script Directory Detection
#!/bin/bash
set -Eeuo pipefail
# Correctly determine script directory
SCRIPT_DIR="$(cd -- "$(dirname -- "${BASH_SOURCE[0]}")" && pwd -P)"
SCRIPT_NAME="$(basename -- "${BASH_SOURCE[0]}")"
echo "Script location: $SCRIPT_DIR/$SCRIPT_NAME"
Pattern 2: Comprehensive Function Templat
#!/bin/bash
set -Eeuo pipefail
# Prefix for functions: handle_*, process_*, check_*, validate_*
# Include documentation and error handling
validate_file() {
local -r file="$1"
local -r message="${2:-File not found: $file}"
if [[ ! -f "$file" ]]; then
echo "ERROR: $message" >&2
return 1
fi
return 0
}
process_files() {
local -r input_dir="$1"
local -r output_dir="$2"
# Validate inputs
[[ -d "$input_dir" ]] || { echo "ERROR: input_dir not a directory" >&2; return 1; }
# Create output directory if needed
mkdir -p "$output_dir" || { echo "ERROR: Cannot create output_dir" >&2; return 1; }
# Process files safely
while IFS= read -r -d '' file; do
echo "Processing: $file"
# Do work
done < <(find "$input_dir" -maxdepth 1 -type f -print0)
return 0
}
Pattern 3: Safe Temporary File Handling
#!/bin/bash
set -Eeuo pipefail
trap 'rm -rf -- "$TMPDIR"' EXIT
# Create temporary directory
TMPDIR=$(mktemp -d) || { echo "ERROR: Failed to create temp directory" >&2; exit 1; }
# Create temporary files in directory
TMPFILE1="$TMPDIR/temp1.txt"
TMPFILE2="$TMPDIR/temp2.txt"
# Use temporary files
touch "$TMPFILE1" "$TMPFILE2"
echo "Temp files created in: $TMPDIR"
Pattern 4: Robust Argument Parsing
#!/bin/bash
set -Eeuo pipefail
# Default values
VERBOSE=false
DRY_RUN=false
OUTPUT_FILE=""
THREADS=4
usage() {
cat <<EOF
Usage: $0 [OPTIONS]
Options:
-v, --verbose Enable verbose output
-d, --dry-run Run without making changes
-o, --output FILE Output file path
-j, --jobs NUM Number of parallel jobs
-h, --help Show this help message
EOF
exit "${1:-0}"
}
# Parse arguments
while [[ $# -gt 0 ]]; do
case "$1" in
-v|--verbose)
VERBOSE=true
shift
;;
-d|--dry-run)
DRY_RUN=true
shift
;;
-o|--output)
OUTPUT_FILE="$2"
shift 2
;;
-j|--jobs)
THREADS="$2"
shift 2
;;
-h|--help)
usage 0
;;
--)
shift
break
;;
*)
echo "ERROR: Unknown option: $1" >&2
usage 1
;;
esac
done
# Validate required arguments
[[ -n "$OUTPUT_FILE" ]] || { echo "ERROR: -o/--output is required" >&2; usage 1; }
Pattern 5: Structured Logging
#!/bin/bash
set -Eeuo pipefail
# Logging functions
log_info() {
echo "[$(date +'%Y-%m-%d %H:%M:%S')] INFO: $*" >&2
}
log_warn() {
echo "[$(date +'%Y-%m-%d %H:%M:%S')] WARN: $*" >&2
}
log_error() {
echo "[$(date +'%Y-%m-%d %H:%M:%S')] ERROR: $*" >&2
}
log_debug() {
if [[ "${DEBUG:-0}" == "1" ]]; then
echo "[$(date +'%Y-%m-%d %H:%M:%S')] DEBUG: $*" >&2
fi
}
# Usage
log_info "Starting script"
log_debug "Debug information"
log_warn "Warning message"
log_error "Error occurred"
Pattern 6: Process Orchestration with Signals
#!/bin/bash
set -Eeuo pipefail
# Track background processes
PIDS=()
cleanup() {
log_info "Shutting down..."
# Terminate all background processes
for pid in "${PIDS[@]}"; do
if kill -0 "$pid" 2>/dev/null; then
kill -TERM "$pid" 2>/dev/null || true
fi
done
# Wait for graceful shutdown
for pid in "${PIDS[@]}"; do
wait "$pid" 2>/dev/null || true
done
}
trap cleanup SIGTERM SIGINT
# Start background tasks
background_task &
PIDS+=($!)
another_task &
PIDS+=($!)
# Wait for all background processes
wait
Pattern 7: Safe File Operations
#!/bin/bash
set -Eeuo pipefail
# Use -i flag to move safely without overwriting
safe_move() {
local -r source="$1"
local -r dest="$2"
if [[ ! -e "$source" ]]; then
echo "ERROR: Source does not exist: $source" >&2
return 1
fi
if [[ -e "$dest" ]]; then
echo "ERROR: Destination already exists: $dest" >&2
return 1
fi
mv "$source" "$dest"
}
# Safe directory cleanup
safe_rmdir() {
local -r dir="$1"
if [[ ! -d "$dir" ]]; then
echo "ERROR: Not a directory: $dir" >&2
return 1
fi
# Use -I flag to prompt before rm (BSD/GNU compatible)
rm -rI -- "$dir"
}
# Atomic file writes
atomic_write() {
local -r target="$1"
local -r tmpfile
tmpfile=$(mktemp) || return 1
# Write to temp file first
cat > "$tmpfile"
# Atomic rename
mv "$tmpfile" "$target"
}
Pattern 8: Idempotent Script Design
#!/bin/bash
set -Eeuo pipefail
# Check if resource already exists
ensure_directory() {
local -r dir="$1"
if [[ -d "$dir" ]]; then
log_info "Directory already exists: $dir"
return 0
fi
mkdir -p "$dir" || {
log_error "Failed to create directory: $dir"
return 1
}
log_info "Created directory: $dir"
}
# Ensure configuration state
ensure_config() {
local -r config_file="$1"
local -r default_value="$2"
if [[ ! -f "$config_file" ]]; then
echo "$default_value" > "$config_file"
log_info "Created config: $config_file"
fi
}
# Rerunning script multiple times should be safe
ensure_directory "/var/cache/myapp"
ensure_config "/etc/myapp/config" "DEBUG=false"
Pattern 9: Safe Command Substitution
#!/bin/bash
set -Eeuo pipefail
# Use $() instead of backticks
name=$(<"$file") # Modern, safe variable assignment from file
output=$(command -v python3) # Get command location safely
# Handle command substitution with error checking
result=$(command -v node) || {
log_error "node command not found"
return 1
}
# For multiple lines
mapfile -t lines < <(grep "pattern" "$file")
# NUL-safe iteration
while IFS= read -r -d '' file; do
echo "Processing: $file"
done < <(find /path -type f -print0)
Pattern 10: Dry-Run Support
#!/bin/bash
set -Eeuo pipefail
DRY_RUN="${DRY_RUN:-false}"
run_cmd() {
if [[ "$DRY_RUN" == "true" ]]; then
echo "[DRY RUN] Would execute: $*"
return 0
fi
"$@"
}
# Usage
run_cmd cp "$source" "$dest"
run_cmd rm "$file"
run_cmd chown "$owner" "$target"
Advanced Defensive Techniques
Named Parameters Pattern
#!/bin/bash
set -Eeuo pipefail
process_data() {
local input_file=""
local output_dir=""
local format="json"
# Parse named parameters
while [[ $# -gt 0 ]]; do
case "$1" in
--input=*)
input_file="${1#*=}"
;;
--output=*)
output_dir="${1#*=}"
;;
--format=*)
format="${1#*=}"
;;
*)
echo "ERROR: Unknown parameter: $1" >&2
return 1
;;
esac
shift
done
# Validate required parameters
[[ -n "$input_file" ]] || { echo "ERROR: --input is required" >&2; return 1; }
[[ -n "$output_dir" ]] || { echo "ERROR: --output is required" >&2; return 1; }
}
Dependency Checking
#!/bin/bash
set -Eeuo pipefail
check_dependencies() {
local -a missing_deps=()
local -a required=("jq" "curl" "git")
for cmd in "${required[@]}"; do
if ! command -v "$cmd" &>/dev/null; then
missing_deps+=("$cmd")
fi
done
if [[ ${#missing_deps[@]} -gt 0 ]]; then
echo "ERROR: Missing required commands: ${missing_deps[*]}" >&2
return 1
fi
}
check_dependencies