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.
git clone --depth 1 https://github.com/synaptiai/synapti-marketplaceWrote 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.
[](https://agentmods.dev/commands/synaptiai/synapti-marketplace/gh-explain)<a href="https://agentmods.dev/commands/synaptiai/synapti-marketplace/gh-explain"><img src="https://agentmods.dev/badge/commands/synaptiai/synapti-marketplace/gh-explain/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/commands/synaptiai/synapti-marketplace/gh-explain"><img src="https://agentmods.dev/badge/commands/synaptiai/synapti-marketplace/gh-explain.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00018 | $0.01858 |
| Opus 5 | $0.00009 | $0.00929 |
| Sonnet 5 | $0.00004 | $0.00372 |
| Haiku 4.5 | $0.00002 | $0.00186 |
Grade B, and why
gh-explain scanned grade B with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured today.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Reads agent configuration directoriesmediumAgent snooping
.claude/, .codex/, .gemini/ hold keys, settings and other credentials a mod has no legitimate need for.
' "$HOME/.claude/settings.gh-workflow.json" \ How it starts
The opening of the file, as written. The whole thing — 197 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Explain: What Did AI Build?
Context loader for interactive Q&A about what AI built on your behalf. Loads the decision journal, diff, issue context, and source files so you can ask questions grounded in evidence.
Tool Usage: This command uses the AskUserQuestion tool for session save decisions.
Contract
GOAL: Load all available context about an issue/PR and present a comprehension brief, then enable conversational Q&A. Testable: user can ask questions and get answers grounded in actual code, decisions, and issue context.
CONSTRAINTS:
- Never fabricate decisions or rationale not present in the journal or diff
- Always distinguish between journal-recorded decisions and inferred decisions
- If context is incomplete, state what's missing rather than guessing
FORMAT: Comprehension brief displayed first, then open-ended conversation.
FAILURE CONDITIONS (output is unacceptable if any apply):
- Claims about decisions that aren't in the journal or diff
- Answers not grounded in loaded context
- Context loading fails silently (must report what's missing)
Phase 1: Context Loading
Step 1.1: Determine Target
If $ARGUMENTS is provided, normalize it to a numeric issue/PR number:
ISSUE_NUM=$(echo "$ARGUMENTS" | grep -oE '[0-9]+$')
If $ARGUMENTS is not provided, detect from current branch:
BRANCH=$(git branch --show-current)
ISSUE_NUM=$(echo "$BRANCH" | grep -oE 'issue-[0-9]+' | grep -oE '[0-9]+')
If no issue number found from branch or arguments, ask via AskUserQuestion tool:
- "Which issue or PR would you like to explore?"
Step 1.2: Load Context (Parallel)
First, read journal directory and explain config from settings:
# Read explain config from settings (local > project > user > default)
GHW_CONFIG=$(jq -n '
def defaults: {
journal: { dir:".decisions" },
explain: { sessionSave:"ask", includeDiff:true }
};
defaults
* (try input catch {})
* (try input catch {})
* (try input catch {})
' "$HOME/.claude/settings.gh-workflow.json" \
".claude/settings.gh-workflow.json" \
".claude/settings.gh-workflow.local.json" 2>/dev/null)
JOURNAL_DIR=$(echo "$GHW_CONFIG" | jq -r '.journal.dir')
INCLUDE_DIFF=$(echo "$GHW_CONFIG" | jq -r '.explain.includeDiff')
SESSION_SAVE=$(echo "$GHW_CONFIG" | jq -r '.explain.sessionSave')
echo "Journal directory: $JOURNAL_DIR"
echo "Include diff: $INCLUDE_DIFF"
echo "Session save: $SESSION_SAVE"
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.
- today First seen · 197 lines · 18 tokens per session scan B 35a892910d7a
gh-explain is a command published in the GitHub repository synaptiai/synapti-marketplace (6 stars, last pushed today), licensed Apache-2.0. It adds 18 tokens to every session and 1,858 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it B with 1 finding (reads agent configuration directories). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-10.
Other commands, from other repositories
cc-review
(Conductor) Review code changes against spec and standards.
create-worktree
Follow these steps to create a git worktree.
production-upgrade
Run the full product upgrade pipeline — 55-agent iterative review with CEO/Engineering/UX/QA parallel loops.
review
Pre-landing code review — analyzes diff for SQL safety, LLM trust boundaries, conditional side effects, missing tests, dependency risks, and security issues.
review
Run code-review and security-review skills against the current diff. Safe to run repeatedly during Build.
checklist
Generate a custom checklist for the current feature based on user requirements.