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 skills add akolotov/harness --skill spawn-review-sessionsgit clone --depth 1 https://github.com/akolotov/harnessWrote 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/skills/akolotov/harness/spawn-review-sessions)<a href="https://agentmods.dev/skills/akolotov/harness/spawn-review-sessions"><img src="https://agentmods.dev/badge/skills/akolotov/harness/spawn-review-sessions/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/skills/akolotov/harness/spawn-review-sessions"><img src="https://agentmods.dev/badge/skills/akolotov/harness/spawn-review-sessions.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 | $0.00037 | $0.01581 |
| Opus 5 | $0.00018 | $0.00790 |
| Sonnet 5 | $0.00007 | $0.00316 |
| Haiku 4.5 | $0.00004 | $0.00158 |
Grade A, and why
spawn-review-sessions 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 9d 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 — 121 lines — stays where its author put it; the contents beside it link to each section on GitHub.
spawn-review-sessions
Thin wrapper over scripts/spawn_review_sessions.py. For each comment block in a
review comments MD file, it seeds a headless Claude Code session and resumes it
under Remote Control inside a backgrounded tmux session, so the operator can
steer each one from claude.ai/code or the mobile app and end it with /exit.
Seeds run in parallel and each session's RC is raised the moment its own seed
finishes (pipelined, no barrier), so the first ready session can be read while
slower seeds are still running.
The input file is the output of the sibling save-review-comments skill.
Path Conventions
Paths in this document use <SKILL_DIR> to mean this skill's installation directory —
for example .claude/skills/spawn-review-sessions in Claude Code, or
.codex/skills/spawn-review-sessions in Codex CLI. Whenever you see <SKILL_DIR> in a
Bash invocation, substitute the actual installation path resolved from the harness. Do
not synthesize an absolute path from your own filesystem assumptions.
<PROJECT_DIR> means the directory the RC sessions start in: --project-dir when
given, otherwise the git root of the current working directory.
Usage
Run the script directly:
python3 <SKILL_DIR>/scripts/spawn_review_sessions.py \
<path/to/comments.md> [--type code-review|plan-review] [--dry-run]
- Review type is auto-detected from the MD file's
meta:blocks (implementation-plan-source-file→ plan-review,github-pr-id→ code-review); override with--type. - Session display names (shown in the RC UI) are derived from
meta::PR#<id> - <slug>for code review,<plan-file-stem> - <slug>for plan review, falling back toreview-<slug>. The tmux handle mirrors this:review-PR<id>-<slug>/review-plan-<ACRONYM>-<slug>(acronym = first letters of the plan-file stem's words), falling back toreview-<slug>. - Artifacts (
seed-<slug>.jsonl/.err,mapping.tsv) are written to asessions/dir next to the MD file. The slug→UUID mapping makes reruns idempotent: live tmux sessions are skipped, previously-seeded slugs resume without re-seeding. - Each seed prompt also tells the session where its own decision report belongs later:
decisions/<slug>.md, next tosessions/(both under the MD file's directory, so a decision report inherits the same run's timestamp instead of a separately-invented one). The prompt says not to write it yet — that happens on a separate, later request within the same session, once a decision has actually been reached. - Phase 1 seeds run with
--output-format stream-json, soseed-<slug>.jsonlfills with events as the model works. The terminalresultevent carries the success/error verdict. - While seeding, the script prints a heartbeat every
progress_interval_seconds(config;0disables): one line per seed with a*per 5 stream lines, ordone/FAILonce finished — so a long-running seed never looks hung. For finer detail,tail -f sessions/seed-<slug>.jsonl | jq -rc .type.
What ships with it
8 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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.
- 9d ago First seen · 121 lines · 37 tokens per session scan A 45fd4d86755d
spawn-review-sessions is a skill published in the GitHub repository akolotov/harness (2 stars, last pushed 9d ago), licensed MIT. It adds 37 tokens to every session and 1,581 once invoked, about $0.0002 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 skills, from other repositories
autoreview
Pre-commit/ship code review: Codex default; optional Claude or Pi.
rework-rate
Measure and interpret PR rework rate — the emerging 5th DORA metric.
omh-code-review
This is a Hermes-native code-review workflow skill.
revdiff-plan
Review the last Codex assistant message (plan, analysis, or proposal) with inline annotations in a TUI overlay. Extracts the most recent response from Codex rollout files and opens it in revdiff for review and annotation. Activates on "revdiff-plan", "review plan with revdiff", "annotate plan", "review last response"…
code-reviewer
Code review specialist focused on patterns, bugs, security, and performance.
agent-teams-simplify-and-harden
Implementation + audit loop using parallel agent teams with structured simplify, harden, and document passes. Spawns implementation agents to do the work, then audit agents to find complexity, security gaps, and spec deviations, then loops until code compiles cleanly, all tests pass, and auditors find zero issues or…