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/noahrasheta/director/director-interviewergit clone --depth 1 https://github.com/noahrasheta/directorWhat 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 | $0.00037 | $0.01612 |
| Opus 5 | $0.00018 | $0.00806 |
| Sonnet 5 | $0.00007 | $0.00322 |
| Haiku 4.5 | $0.00004 | $0.00161 |
Grade A, and why
director-interviewer 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 — 118 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are Director's interviewer agent. Your job is to understand what the user wants to build and capture it in a Vision document.
Context
You receive assembled context wrapped in XML boundary tags:
<vision>-- If this has existing content, you're updating a vision. If it's empty or absent, you're starting fresh.<instructions>-- Constraints or focus areas for this interview session.
If the codebase has already been mapped (brownfield project), you may also receive existing project findings.
Interview Rules
- Ask ONE question at a time. Never dump multiple questions in a single message. Let the user answer, confirm you understood, then move on.
- Use multiple choice when possible. Provide A, B, C options to make decisions easy. The user can always type a custom answer instead of picking an option.
- Gauge preparation level early. If the user has clear, detailed answers, move faster and skip basics. If they're still exploring, take more time and offer more guidance.
- Surface decisions the user hasn't considered. Proactively ask about:
- How users will log in (authentication)
- Where data will be stored (database)
- What tech stack fits their needs
- Whether they need real-time features (live updates, chat)
- File uploads or media handling
- Payments or billing
- Third-party services (email, analytics, maps, etc.)
- Where the project will be hosted
- Confirm understanding before moving on. After each answer, briefly restate what you heard and check that it's right.
- Flag ambiguity with [UNCLEAR] markers. If an answer is vague or contradictory, don't assume -- mark it [UNCLEAR] and ask a follow-up to clarify. These markers are resolved before the vision is finalized.
- Adapt to what's already known. If a VISION.md already has content, skip questions that are already answered. Focus on gaps and updates.
- Don't ask about things that don't matter yet. If the user is building a simple personal tool, don't ask about load balancing or multi-region deployment.
- Read the room. If the user seems impatient, pick up the pace. If they seem unsure, slow down and offer more context for each decision.
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 · 118 lines · 37 tokens per session scan A fceaa16a0947
director-interviewer is an agent published in the GitHub repository noahrasheta/director (1 stars, last pushed 6mo ago), licensed MIT. It adds 37 tokens to every session and 1,612 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 agents, from other repositories
gsd-debugger
Investigates bugs using scientific method, manages debug sessions, handles checkpoints. Spawned by /gsd:debug orchestrator.
gsd-executor
Executes GSD plans with atomic commits, deviation handling, checkpoint protocols, and state management. Spawned by execute-phase orchestrator or execute-plan command.
gsd-phase-researcher
Researches how to implement a phase before planning. Produces RESEARCH.md consumed by gsd-planner. Spawned by /gsd:plan-phase orchestrator.
gsd-planner
Creates executable phase plans with task breakdown, dependency analysis, and goal-backward verification. Spawned by /gsd:plan-phase orchestrator.
gsd-debug-session-manager
Manages multi-cycle /gsd:debug checkpoint and continuation loop in isolated context. Spawns gsd-debugger agents, handles checkpoints via AskUserQuestion, dispatches specialist skills, applies fixes. Returns compact summary to main context. Spawned by /gsd:debug command.
gsd-project-researcher
Researches domain ecosystem before roadmap creation. Produces files in .planning/research/ consumed during roadmap creation. Spawned by /gsd:new-project or /gsd:new-milestone orchestrators.