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 commands/jeet129/praxis/reviewgit clone --depth 1 https://github.com/jeet129/praxisWrote 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/jeet129/praxis/review)<a href="https://agentmods.dev/commands/jeet129/praxis/review"><img src="https://agentmods.dev/badge/commands/jeet129/praxis/review.svg" alt="Measured on agentmods" 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 | $0.00055 | $0.01237 |
| Opus 5 | $0.00028 | $0.00619 |
| Sonnet 5 | $0.00011 | $0.00247 |
| Haiku 4.5 | $0.00006 | $0.00124 |
Grade A, and why
review 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 5d 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 — 50 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Review the artifact named in $ARGUMENTS (e.g. contracts, roadmap, adrs, or a file/dir path).
This command exists for architecture-significant artifacts produced outside a governance gate — e.g. a contract baseline authored after architecture_sign_off, or a roadmap revised mid-build. It runs a structured, multi-role review AND closes the loop: findings are routed back to the responsible agent, fixed, and re-reviewed until no BLOCK remains. It does NOT replace architecture_sign_off or the per-slice review gates; it produces findings, drives remediation, and leaves a record.
Step 1 — Resolve scope from $ARGUMENTS:
contracts→contracts/(schemas, openapi, proto, events). Load the checklistapi-design/references/contract-review.md.roadmap→ the phased-roadmap section in.project/working/architecture/.adrs→.project/decision/(the ADR set + INDEX).<path>→ that specific file or directory. Read the artifact plus its governing inputs:.project/semantic/project-charter.md, the relevant ADRs,nfr-register.md, and the shared conventions (M00 /contracts/schemas/).
Step 2 — Run reviewers in parallel, each with its lens:
- architecture-challenger — design soundness: coupling, evolvability/versioning, the cross-tier seam (BE↔FE↔AI), behavior under the NFR budgets. Use the relevant sub-personas.
- code-reviewer (Dimension 6: API + data contracts) — schema hygiene,
$refconsistency, backward-compatible-only or explicit version bump, expand-contract, single error-model usage, naming. - security-reviewer — safety/privacy surfaces: prohibited-intent generative paths ABSENT (per the project's safety ADRs), authz on every endpoint, server-bound actor, idempotency keys, PII at boundaries, tenant isolation.
For
contracts, every reviewer usesapi-design/references/contract-review.md.
Step 3 — Consolidate findings. For each: classify BLOCK / FIX / ACCEPT, map it to the ADR/NFR/convention it touches, and assign an owner (the agent responsible for the fix) per the ownership map:
| Artifact / finding type | Owner agent | Skill it fixes with |
|---|---|---|
| Contract surface, API/data shape, versioning | solution-architect | api-design, data-modeling |
| Phased roadmap / slicing | solution-architect | project-phasing |
| ADRs / decisions | solution-architect (principal for the decision) | adr-decision-records |
| Safety/privacy on AI surfaces (M17/M18/M19) | solution-architect + ml-ai-engineer | llm-safety, agentic-architecture |
| Per-module endpoint detail (when its slice lands) | the specialist owning that slice | the relevant stack-* |
Step 4 — Route and remediate (the loop). This is the point of the command — do not stop at a findings list:
- Dispatch the BLOCK + FIX findings to the owning agent(s) from the map. Give each owner the specific finding, the ADR/NFR it touches, and the expected fix.
- The owner fixes the artifact (SA edits the contract/roadmap/ADR; ml-ai-engineer fixes AI-safety surfaces; etc.).
- Re-review the changed surface (re-run the relevant reviewer on just the touched files).
- Repeat until no BLOCK findings remain. FIX items may be closed now or tracked with an owner + due slice.
Step 5 — Disposition of disputes / accepted risk. If an owner disagrees with a finding or wants to accept it rather than fix it, do not silently drop it — route it to the principal (mirrors challenger_objection_override): the principal accepts (with rationale, recorded as an ADR) or rejects (back to remediation). Safety/privacy BLOCKs (Step 2 security lens) are not accept-able by the owner alone; they require explicit principal sign-off.
Step 6 — Record + status. Write the review record to .project/operational/reviews/<artifact>-<YYYY-MM-DD>.md with the findings table (id · severity · surface · ADR/NFR · owner · disposition · status) and a top-line status: OPEN → IN-REMEDIATION → CLOSED. Update .project/working/active-workflow.md. A review is CLOSED only when every BLOCK is fixed-and-re-reviewed or principal-accepted.
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.
- 5d ago First seen · 50 lines · 55 tokens per session scan A df77948e9d62
review is a command published in the GitHub repository jeet129/praxis (7 stars, last pushed yesterday), licensed MIT. It adds 55 tokens to every session and 1,237 once invoked, about $0.0003 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 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.
analyze
Perform a non-destructive cross-artifact consistency and quality analysis across spec.md, plan.md, and tasks.md after task generation.
constitution
Create or update the project constitution from interactive or provided principle inputs.
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.