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/hffmnnj/goopspec/goop-executegit clone --depth 1 https://github.com/hffmnnj/goopspecWhat 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 | $0.00013 | $0.01309 |
| Opus 5 | $0.00006 | $0.00655 |
| Sonnet 5 | $0.00003 | $0.00262 |
| Haiku 4.5 | $0.00001 | $0.00131 |
Grade A, and why
goop-execute 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 — 79 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/goop-execute
Implement the blueprint one wave at a time. The orchestrator delegates all implementation to executor agents.
Gate check
Call goop_state({ action: "get" }). If specLocked is not true or no wave rows exist — check via goop_read_wave({ workflow_id }) returning at least one wave (not a 'no waves' message) — return BLOCKED with:
Run
/goop-planfirst.
Load references
goop_reference({ name: "dispatch-patterns" })
goop_reference({ name: "core-protocol" })
Steps
- Read
goop_read_db({ doc_types: ["spec", "blueprint", "chronicle"] })for narrative context andgoop_read_wave({ workflow_id })for wave/task context; readPROJECT_KNOWLEDGE_BASE.md. - For each wave:
- Delegate tasks to the right executor tier via
task():goop-executor-lowfor mechanical/pattern-following edits; escalate if hidden complexity.goop-executor-mediumfor standard implementation work — the default tier when a task isn't clearly mechanical or complex/critical.goop-executor-highfor architecture-sensitive or security-critical work — use when genuinely warranted.goop-executor-frontend-lowfor UI mechanical tasks; escalate if hidden complexity.goop-executor-frontend-mediumfor standard UI component work — the default frontend tier when UI work isn't clearly mechanical or design/architectural.goop-executor-frontend-highfor deep design or UI-architecture work — use when genuinely warranted.
- Use sequential dispatch for shared files; parallel dispatch for independent tasks.
- Require every task to return
STATUS,SUMMARY,ARTIFACTS,VERIFICATION,NEXT.
- Delegate tasks to the right executor tier via
- Apply the four-rule deviation system from
phase-gates. Log every deviation toADL.mdviagoop_adl. - Before dispatching a wave, call
goop_state({ action: "update-wave", currentWave: N, totalWaves: M })to record that Wave N is in progress.currentWaveis 1-based;0means no wave has started. - After every task in a wave completes, run the wave verification gate before the wave can be marked complete:
- Dispatch
goop-wave-verifierscoped to that wave — one wave per dispatch; the role is execute-phase and inspect-only (seeagents/goop-wave-verifier.md). - The verifier records a verification row per task via
goop_write_wave({ wave_number: N, verifications: [...] })and never implements fixes. - On gaps: dispatch the correctly tiered executor to remediate, then re-dispatch
goop-wave-verifierfor that wave to re-verify the remediated tasks. - Bound the remediation cycle with the three-strikes convention in
references/test-authoring.md§Test Execution Discipline: after three failed attempts on the same gap, open a blocker viagoop_blockerand stop; do not loop. - The wave is complete when its rows satisfy the runtime gate: verification rows persist forever (never deleted or edited), and the gate reads the latest row per
check_name— at least one check present, and every check's latest row non-fail. A laterpassor explicitskipfor the same check supersedes that check's own priorfail; it does not resolve a different check's unresolved fail (isWaveVerifiedinsrc/features/enforcement/verifier-stage.ts). The runtime enforces the boundary:goop_write_waverejects a complete-status transition without it.
- Dispatch
- If Atomic PRs = Yes: after the wave verification gate passes, open a PR for the verified wave against the previous branch (Wave N → Wave N-1; Wave 1 → main) via
gh pr createorgoop_create_pr. Do not wait for it to merge. Then create the Wave N+1 branch from the current wave's branch and continue. Show the PR URL in the checkpoint. - Save a checkpoint at wave boundaries.
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 · 79 lines · 13 tokens per session scan A 470b012f31c1
goop-execute is a command published in the GitHub repository hffmnnj/goopspec (46 stars, last pushed 20d ago), licensed MIT. It adds 13 tokens to every session and 1,309 once invoked, about $0.0001 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 2026-08-30.
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.