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/vrennat/developersdevelopers/implgit clone --depth 1 https://github.com/vrennat/developersDevelopersWhat 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 | $0.00040 | $0.01467 |
| Opus 5 | $0.00020 | $0.00733 |
| Sonnet 5 | $0.00008 | $0.00293 |
| Haiku 4.5 | $0.00004 | $0.00147 |
Grade A, and why
impl 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 yesterday.
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 — 76 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/impl
Execute work. Input is one of:
- Path to spec file:
/impl docs/specs/foo.md - Ticket ID:
/impl ERT-1234(uses Linear MCP if available) - Freeform:
/impl "make banner sticky on mobile" --dry-runflag: print the assessment and stop
Procedure
-
Parse input. If it matches
[A-Z]+-\d+and Linear MCP tools (mcp__plugin_linear_linear__*ormcp__claude_ai_Linear__*) are available, fetch the ticket. If it's a path, read the spec. Otherwise treat as freeform. -
Read mode, then classify clarity, complexity, AND stakes.
- Mode: read
.claude/dd-mode(treat absent asdefault). It iscautious,default, orautonomous— set via/mode. - Clarity: clear (one obvious approach) or ambiguous (2+ approaches with real tradeoffs, OR missing requirement, OR multi-cause bug).
- Complexity: simple (1 file, <50 LOC), medium (2-3 files), complex (>3 files).
- Stakes: high if it touches auth, money, data integrity, security, or privacy, or is hard to undo; otherwise normal.
- Print the mode first, then the classification, as the top two lines of output:
Mode: <mode>Clarity: clear/ambiguous | Complexity: simple/medium/complex | Stakes: normal/high
- Mode: read
-
Branch on clarity, modulated by mode.
default: ambiguous → ask all open questions in ONE batched numbered list, wait, then proceed. Clear → proceed silently.cautious: same as default, plus — before executing a medium or complex task even when clarity is clear, print the planned routing and wait for a go-ahead.autonomous: never block on clarity. Ambiguous → pick the most sensible default, state the assumption in one line, and proceed. (Destructive/irreversible ops and stakes-gated review are unaffected by mode — see Rules.)
-
If
--dry-run: print the planned routing and stop. -
If Linear ticket: update status to "In Progress" via Linear MCP.
-
Second look (medium/complex only — skip for simple). Before writing any code, interrogate the direction you just picked, in one deliberate pass. The first approach that looks right is usually the statistically-likely default, not the considered one. Ask: what here is the generic pattern reached for by reflex? What would give this a point of view instead of "looks fine"? What can be cut or tightened? Is there a simpler path dismissed too quickly? One pass, then commit to a direction — a sharpening step, not a stalling loop.
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.
- yesterday First seen · 76 lines · 40 tokens per session scan A 595c1e5cafd2
impl is a command published in the GitHub repository vrennat/developersDevelopers (2 stars, last pushed 1mo ago), licensed MIT. It adds 40 tokens to every session and 1,467 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 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.
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.