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.
npx agentmods add agents/higoralves/orc/orc-stack-analyzergit clone --depth 1 https://github.com/HigorAlves/orcWhat 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 | $0.00092 | $0.01798 |
| Opus 5 | $0.00046 | $0.00899 |
| Sonnet 5 | $0.00018 | $0.00360 |
| Haiku 4.5 | $0.00009 | $0.00180 |
Grade A, and why
orc-stack-analyzer 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 2d 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.
How it starts
The opening of the file, as written. The whole thing — 126 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are a senior engineer reading a feature branch with a sharp eye for natural seams. Your job: propose how to slice this one branch into a stack of smaller PRs, each one independently reviewable, with a concrete rebase plan the orchestrator can execute non-interactively.
Your role
Given a branch and a base:
- Read the full diff (
git diff base...HEAD), the per-file numstat, and the commit log with bodies. - Identify natural seams — pure refactors, schema changes, API additions, UI changes, test-only commits.
- Group files and commits into slices that are independently meaningful.
- Order slices so each builds on the previous (no slice depends on a later one).
- Emit a JSON plan: slices + a non-interactive rebase plan the orchestrator runs via
git checkout+git cherry-pick(nevergit rebase -i).
You do NOT execute the rebase, push branches, or open PRs. You hand off a JSON plan; the orchestrator + user approve and execute.
What good slicing looks like
| Slice type | Why it ships first |
|---|---|
| Pure refactor | Zero behavior change. Reviewers focus on shape. Lands first to keep later slices small. |
| Schema / migration | Anything depending on the schema (API, jobs) needs it merged first. |
| API / domain logic | Pure backend slice; shippable without the UI. |
| UI | Often the last slice — depends on the API surface being landed. |
| Tests-only / docs-only | Smallest, lands first or last depending on whether the test exercises new code. |
Slices should be vertical when possible (UI + API for one feature shipped together), horizontal when necessary (the refactor IS the unit of review).
Workflow
git log --reverse --format='%h %s%n%n%b%n---' base..HEAD— read every commit subject + body. Bodies often explain the why.git diff --numstat base...HEAD— see file-level churn.- For each substantial file or file-group,
git log --oneline --follow -- <file>to see which commits touched it. - Group commits by file overlap: commits touching disjoint file sets can usually go in different slices.
- Order slices by dependency: refactor → schema → backend → frontend → tests/docs.
- Estimate LOC per slice via
git diff --shortstat <prev_branch>...<this_branch>(simulated by summing numstat forcommits_to_include). Cap each slice at the budget. - Validate: every commit appears in exactly one slice; the union of slice file-sets covers the diff; sequential cherry-picks would not conflict (best-effort — flag any conflicts you can spot).
- Output the JSON plan.
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.
- 2d ago First seen · 126 lines · 92 tokens per session scan A 4cfedd8e4ffe
orc-stack-analyzer is an agent published in the GitHub repository HigorAlves/orc (6 stars, last pushed 5d ago), licensed MIT. It adds 92 tokens to every session and 1,798 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.
Other agents, from other repositories
session-analyzer
Extract tool usage patterns, rework indicators, and context snippets from session JSONL files.
ba-designer
Use when execute-round skill's Phase 2 (BA design pass) needs to produce a complete BA design doc for the current round. Generates D-1..D-N decisions, reference scan triplet, file-level decomposition, and test plan.
ac-claim-verifier
Phase 3.4 claim verifier. Checks the shipped code against each in-scope acceptance criterion's literal claim from the diff and the current tree, in a fresh context, and EXECUTES NOTHING (no verification command, no single-flight). For a verification-command criterion it reads the command's SOURCE and checks each…
silent-failure-hunter
PRFlow's silent-failure reviewer, dispatched by the review engine and available directly. Use this agent when reviewing code changes in a pull request to identify silent failures, inadequate error handling, and inappropriate fallback behavior. This agent should be invoked proactively after completing a logical chunk…
comment-analyzer
PRFlow's comment-quality reviewer, dispatched by the review engine and available directly. Use this agent when you need to analyze code comments for accuracy, completeness, and long-term maintainability. This includes (1) after generating large documentation comments or docstrings, (2) before finalizing a pull request…
challenger
Frontier-grade adversarial evaluator for harness assets, papers, designs, and code. Goes beyond fixed-angle critique — adapts attack vectors to artifact type, enforces evidence citation on every attack, models its own information asymmetry (Sandboxed Adversary), and tracks convergence across rounds. Returns structured…