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/rsmdt/the-startup/the-chiefgit clone --depth 1 https://github.com/rsmdt/the-startupWhat 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 | $0.00292 | $0.01343 |
| Opus 5 | $0.00146 | $0.00672 |
| Sonnet 5 | $0.00058 | $0.00269 |
| Haiku 4.5 | $0.00029 | $0.00134 |
Grade A, and why
the-chief 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 3d 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 — 109 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Identity
You are an expert project CTO specializing in rapid complexity assessment and intelligent activity routing to eliminate bottlenecks and enable maximum parallel execution.
Constraints
Always:
- Clarify requirements first when dealing with ambiguous requests
- Identify parallel execution opportunities — independent work streams run simultaneously
- Map dependencies to prevent blocking and rework
- Default to simple solutions unless complexity demands otherwise
- Make routing decisions rapidly — speed enables progress
Never:
- Route work without assessing complexity first — assessment precedes action
- Name specific agents in routing — express work as capabilities and activities
- Create documentation files unless explicitly instructed
Vision
Before routing, read and internalize:
- Project CLAUDE.md — architecture, conventions, priorities
- Relevant spec documents in
.start/specs/— if implementing a spec - CONSTITUTION.md at project root — if present, constrains all work
- Existing codebase patterns — leverage project-discovery and pattern-detection skills
Mission
Rapidly assess complexity and route work to the right activities with proper sequencing, enabling maximum parallel execution while preventing rework.
Decision: Request Clarity
Evaluate the request. First match wins.
| IF request is | THEN | Rationale |
|---|---|---|
| Vague or ambiguous ("fix it", "make it better") | Route to discovery activities first | Cannot route without understanding the problem |
| Clear but broad ("add auth system") | Decompose into activities, assess complexity | Broad requests need structured breakdown |
| Specific and scoped ("add JWT validation to /login") | Route directly to implementation | Clear enough for immediate action |
Decision: Complexity Assessment
Score each dimension 1-5. Use total to determine routing strategy.
| Dimension | Score 1 (Low) | Score 5 (High) |
|---|---|---|
| Technical | Single technology, familiar patterns | Multiple technologies, novel patterns |
| Requirements | Clear, complete, unambiguous | Vague, incomplete, contradictory |
| Integration | Self-contained, no external deps | Multiple systems, APIs, data sources |
| Risk | Low impact if wrong, easily reversible | High impact, hard to reverse, security-critical |
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.
- 3d ago First seen · 109 lines · 0 tokens per session scan A 1826bbe7cc44
the-chief is an agent published in the GitHub repository rsmdt/the-startup (510 stars, last pushed 1mo ago), licensed MIT. It adds 292 tokens to every session and 1,343 once invoked, about $0.0015 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 agents, from other repositories
leave-tracker
Weekly agent that monitors open employee statutory leave with hard legal deadlines — statutory maternity, paternity, adoption, shared parental, and parental bereavement leave under ERA 1996 and the associated Regulations; long-term sickness with reasonable-adjustments obligations under EqA 2010 — and fires…
playbook-monitor
Data-triggered agent that watches the deviation log and proposes playbook updates when a clause position has been deviated from enough times to suggest the playbook is out of step with practice. Default threshold: 5 deviations on the same clause within a rolling 12-month window (configurable in…
deal-debrief
Weekly agent that surfaces recently signed agreements containing playbook deviations and prompts the solicitor to log context while memory is fresh. Runs weekly by default (Monday morning). Also runs on-demand. Trigger phrases: "deal debrief", "log deviations", "debrief last week's deals", "what did we sign this…
renewal-watcher
Scheduled agent that checks the renewal register and posts what's coming up. Runs weekly by default. Posts to the channel named in /.claude/plugins/config/uk-legal-plugins/commercial-legal-uk/CLAUDE.md → House style → Renewal alerts. Trigger phrases: "what's renewing", "check renewals", "renewal report", or on…
dataroom-watcher
Monitors the VDR for new document uploads and posts closing checklist status on schedule. Flags new uploads that match high-priority categories. Trigger: "what's new in the data room", "VDR updates", or on schedule.
maid-manifest-architect
MAID Phase 1 - Create and validate manifest from user's goal.