Getting it into your agent
It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.
git clone --depth 1 https://github.com/mathiasbourgoin/rosternpx agentmods add commands/mathiasbourgoin/roster/roster-configWrote 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/mathiasbourgoin/roster/roster-config)<a href="https://agentmods.dev/commands/mathiasbourgoin/roster/roster-config"><img src="https://agentmods.dev/badge/commands/mathiasbourgoin/roster/roster-config/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/mathiasbourgoin/roster/roster-config"><img src="https://agentmods.dev/badge/commands/mathiasbourgoin/roster/roster-config.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 | $0.00015 | $0.01276 |
| Opus 5.5 | $0.00006 | $0.00510 |
| Sonnet 5.5 | $0.00003 | $0.00255 |
| Haiku 4.5 | $0.00002 | $0.00128 |
Grade A, and why
roster-config 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 1mo 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 — 127 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Roster Config
You help users discover and configure the tunables exposed by their installed roster agents. You read installed agent files, surface available settings grouped by impact tier, and rewrite frontmatter for any changes the user approves.
Scope: operates on agents installed in the current project (.claude/commands/ or .harness/agents/ or agents/). Falls back to the roster source tree if no installed harness is found.
Steps
Step 1 — Locate installed agents
# Priority order: installed harness → claude commands → roster source
[ -d .harness/agents ] && echo ".harness/agents" || \
([ -d .claude/commands ] && echo ".claude/commands") || \
([ -d agents ] && echo "agents (roster source)")
If no agent directory is found, report:
⚠️ No installed harness found. Run
/roster-initfirst, or run this from a project with an installed roster harness.
Step 2 — Collect all tunables
Walk all .md files in the found directory. For each file with a tunables: frontmatter block, extract:
- Agent name (from
name:or filename) - Each key-value pair under
tunables: - Inline comment if present (the
# ...after the value)
Build an inventory: { agent, key, value, description, has_ext_dep }.
Mark has_ext_dep: true for any tunable whose inline comment or description mentions pip install, npm install, brew install, or a package name.
Step 3 — Display grouped inventory
Present tunables grouped by tier:
📦 Roster Config — Installed Tunables
🔴 Tier 1 — Optional capabilities with external deps
kb-agent › search_index false # set true to enable LanceDB semantic index
dep: pip install lancedb OR npm install @lancedb/lancedb
🟡 Tier 2 — Project-specific paths (must be customized)
migration-guard › schema_module "" # path to module containing current_version and all_ddl
migration-guard › migration_test_file "" # path to the migration-path test file
pr-workflow › pre_push_validator "" # e.g. "gitwright validate" — leave blank to skip
pr-workflow › pre_pr_checks "" # project-specific pre-PR command
🟢 Tier 3 — Feature toggles (safe defaults, may want to flip)
kb-agent › embedding_mode remote # remote | disabled; disabled keeps deterministic gates offline
implementer › use_worktree true
reviewer › require_security_pass true
tech-lead › require_review true
tech-lead › require_qa true
architect › enforce_architecture_doc true
...
⚙️ Tier 4 — Numeric thresholds
architect › max_file_lines 500
architect › max_function_lines 50
tech-lead › max_parallel_implementers 5
...
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.
- 1mo ago First seen · 127 lines · 15 tokens per session scan A ed04fa06ea66
roster-config is a command published in the GitHub repository mathiasbourgoin/roster (2 stars, last pushed 2d ago), licensed MIT. It adds 15 tokens to every session and 1,276 once invoked, about $0.0001 per session on Opus 5.5. 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 2026-09-07.
Other commands, from other repositories
checklist
Generate a custom checklist for the current feature based on user requirements.
clarify
Identify underspecified areas in the current feature spec by asking up to 5 highly targeted clarification questions and encoding answers back into the spec.
specify
Create or update the feature specification from a natural language feature description.
converge
Assess the current codebase against the feature's spec, plan, and tasks, then append any remaining unbuilt work as new tasks to tasks.md so implement can complete it.
implement
Execute the implementation plan by processing and executing all tasks defined in tasks.md.
speckit.bug.fix
Apply the remediation from a bug assessment and record what was changed.