impl

A command that builds an application in visible stages: first the API, then the web interface, with a user check between them.

In plain words
What is it for?
Use it after an API design contract exists to create the server and web code, then prepare the project for verification.
Why use it?
It replaces hidden parallel work with a traceable sequence, so progress and decisions are visible and work time is easier to review.

Command

Part of the digital-delivery-team plugin — 13 skills, 21 commands, 9 agents, 8 hooks shipped together

Install

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.

agentmods
npx agentmods add commands/dhslegen/digital-delivery-team/impl
Clone the repo
git clone --depth 1 https://github.com/dhslegen/digital-delivery-team

Or install digital-delivery-team, the plugin that ships this one along with the rest of its 13 skills, 21 commands, 9 agents, 8 hooks.

Per session 45 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,727 The whole file, excluding the scripts and references it only reads on demand.
Security scan B 1 finding. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5 $0.00045 $0.01727
Opus 5 $0.00023 $0.00864
Sonnet 5 $0.00009 $0.00345
Haiku 4.5 $0.00005 $0.00173

Measured 2d ago against content hash 94d4135c38a7, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade B, and why

impl scanned grade B with 1 finding 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.

Reads agent configuration directoriesmediumAgent snooping

.claude/, .codex/, .gemini/ hold keys, settings and other credentials a mod has no legitimate need for.

[ -f "$DDT_PLUGIN_ROOT/bin/aggregate.mjs" ] || DDT_PLUGIN_ROOT=$(cat "${HOME}/.claude/delivery-metrics/.ddt-plugin-root" 2>/dev/null)
commands/impl.md · 171 lines

How it starts

The opening of the file, as written. The whole thing — 171 lines — stays where its author put it; the contents beside it link to each section on GitHub.

/impl

M6.4 重大变更:v0.6.x 用 frontend-agent + backend-agent 同消息并行派发(黑盒);v0.7.0 改为串行调用 /build-api/build-web,每步 main thread 流式可见,每步用户决策门。

彻底解决:

  • 工时不可证明(lookback join 并发错配)
  • 黑盒(subagent 黑盒派发,用户看不到)
  • 用户失语(盲盒严重)

输入:$ARGUMENTS


Phase 1 — 前置校验

test -f docs/api-contract.yaml || { echo "❌ 请先运行 /design"; exit 1; }
[ -f "$DDT_PLUGIN_ROOT/bin/aggregate.mjs" ] || DDT_PLUGIN_ROOT=$(cat "${HOME}/.claude/delivery-metrics/.ddt-plugin-root" 2>/dev/null)
[ -f "$DDT_PLUGIN_ROOT/bin/aggregate.mjs" ] || DDT_PLUGIN_ROOT="${HOME}/.claude/plugins/marketplaces/digital-delivery-team"
[ -f "$DDT_PLUGIN_ROOT/bin/aggregate.mjs" ] || { echo "❌ DDT plugin root 未解析"; exit 1; }
export DDT_PLUGIN_ROOT
if printf '%s' "$ARGUMENTS" | grep -q -- '--dry-run'; then
  node "$DDT_PLUGIN_ROOT/bin/print-dry-run.mjs" --phase impl --inputs "docs/api-contract.yaml,.ddt/tech-stack.json,docs/design-brief.md (spa)" --outputs "server/,web/" --next "/verify" --notes "编排命令: 串行调用 /build-api → 决策门 → /build-web"
  exit 0
fi
node "$DDT_PLUGIN_ROOT/bin/emit-phase.mjs" --phase impl --action start
"$DDT_PLUGIN_ROOT/bin/check-blockers.sh" || exit 2

# 契约 lint 硬门禁
if command -v npx >/dev/null 2>&1; then
  npx --yes @redocly/cli lint docs/api-contract.yaml || exit 4
else
  echo "OpenAPI lint tool missing"; exit 5
fi

串行执行流程

Step 1: 后端实现(除非 --web-only)

if ! printf '%s' "$ARGUMENTS" | grep -q -- '--web-only'; then
  echo "▶ 进入后端实现:/build-api"
  # 透传 --module / --auto / --refresh
fi

main thread 调用 /build-api $ARGUMENTS_FILTERED(过滤掉 --web-only)。

/build-api 内部走 6-phase(EXPLORE → PLAN → APPROVE → IMPLEMENT → VERIFY → SUMMARY)+ 决策门:

  • 用户在 PLAN 阶段批准 plan
  • IMPLEMENT 每 step validation + checkpoint
  • SUMMARY 后决策门 4 选项

未传 --auto 时 main thread 必须等待 build-api 决策门返回 accept 才进入 Step 2

Step 2: 前端实现(除非 --api-only)

if ! printf '%s' "$ARGUMENTS" | grep -q -- '--api-only'; then
  echo "▶ 进入前端实现:/build-web"
fi

main thread 调用 /build-web $ARGUMENTS_FILTERED(过滤掉 --api-only)。

Read the full file on GitHub · 171 lines

Changes

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.

  1. 2d ago First seen · 171 lines · 45 tokens per session scan B 94d4135c38a7

Subscribe to this mod's changes

impl is a command published in the GitHub repository dhslegen/digital-delivery-team (1 stars, last pushed 3mo ago), licensed MIT. It adds 45 tokens to every session and 1,727 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it B with 1 finding (reads agent configuration directories). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.