agentleFS
datadog-labsdatadog-labs/agent-skills/dd-orchestrator/SKILL.md

dd-orchestrator

Skill185 starsChanged 38 days agoMITView on GitHub

Entry point for Datadog onboarding. Takes a developer's plain-language goal, ensures a valid Datadog account with dd-account-setup, asks dd-product-recommender which products fit, detects the project's platform and cloud, then composes an ordered plan across the existing skills (agent install, product enable, verify, and optional cloud integration) and dispatches to each by source URL — honestly flagging products with no skill yet. Use when the user says "set up Datadog", "onboard my app / this repo to Datadog", "instrument my project", or states a monitoring goal without naming a specific product or skill.

  • Reads credentials

The file

Show its rules and text

We split this skill into the rules it gives an agent

20 rules. About 0 are common, 8 are distinctive, and 12 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈8

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 12

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    No plan, no dispatch: If you have not run resolve.py and captured its PLAN block (with the SESSION ID) for this run, dispatch nothing — the …

    line 57
  2. Distinctive:

    Do not hand-build a plan.

    line 59
  3. Very specific:

    Every id must be in the PLAN: Before you dispatch a skill id, confirm it appears in that PLAN block (e.g.

    line 60
  4. Very specific:

    If the skill id is in the session registry, invoke it directly.

    line 66
  5. Very specific:

    Sequential dispatch — chains first, stop on failure: Run the plan strictly in resolve.py's order (it is deterministic and independent of …

    line 79
  6. Distinctive:

    All user-facing output — the checklist, the plan, choice questions, gaps, and the summary — uses Simplified Technical English (ASD-STE100).

    line 102
  7. Distinctive:

    Keep sentences short: at most 20 words for an instruction, 25 for a description.

    line 104
  8. Very specific:

    Use the active voice, the imperative for instructions, and simple verb tenses.

    line 106
  9. Distinctive:

    Keep the articles ("a", "the") and write complete sentences.

    line 108
  10. Distinctive:

    Use one word for one meaning, and keep the same term for the same thing.

    line 109
  11. Distinctive:

    Prefer short, common words.

    line 111
  12. Very specific:

    If it prints one or more product tokens, the user named the products: use that list and skip the recommender.

    line 145
  13. Very specific:

    If it prints nothing, the intent describes a goal: you MUST invoke dd-product-recommender and use its ranked product list (installed → …

    line 147
  14. Very specific:

    Write the output to a new file and replace $TRACE only when the command succeeds, so a failed re-run cannot empty the trace:

    line 198
  15. Distinctive:

    Confirm — render the PLAN block, then dispatch: Before any dispatch, render the PLAN block (see *Output templates* below) verbatim: fill …

    line 201
  16. Very specific:

    After record declined, go to Step 8: record run-finished closes every unrun node as declined.

    line 222
  17. Distinctive:

    Rendering rules (both blocks): Use the given section order and the exact headers.

    line 239
  18. Very specific:

    When the trace has a CAVEATS: block, check each caveat against what the developer asked for and the detected context before the confirm …

    line 284
  19. Very specific:

    Every dispatch boundary is one emit.py record call.

    line 339
  20. Very specific:

    Reliability and reconciliation: Every event carries a per-session event_seq (a monotonic ordinal): a gap in event_seq means an event was …

    line 385

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Datadog Onboarding Orchestrator

You are the entry point for Datadog onboarding — the conductor, not a performer. You take one plain-language goal, decide which existing skills are needed and in what order, and hand off to them across all sources. You do not write instrumentation yourself; each skill owns its steps.

Routing is compositional, not a lookup. A product like "APM" is not one skill — it expands into ensure account → install the Agent for the detected platform → enable the product → verify, and surfaces any relevant cloud integration as an optional suggestion (never a required step). That composition is computed from the capability graph in catalog.json; there is no intent-to-skill table anywhere (intents live only in the recommender).

Ground rules (read once)

  • Confirm before you run. Show the composed plan (skills, in order) and the skipped dead-ends, and get a yes before dispatching anything — with one exception: dd-account-setup runs first as a preflight (Step 2), because there is no plan to show until an account exists. Nothing else dispatches before approval.
  • The catalog is the source of truth. catalog.json holds every skill as a node with facets (kind, product, platform, cloud), a category-level requires graph, optional after ordering behind products already in the plan, and a source.url. resolve.py composes the plan from it. Do not hand-maintain a routing table.
  • Only enabled skills route. Each node carries enabled (the two public sets — agent-skills
    • dd-source status: ga — plus this repo's own skills; everything else is disabled). resolve.py composes only from enabled skills in real time; a disabled skill is treated as unavailable and surfaces as a dead-end (demand signal).
  • Use repository evidence and the developer's stated context. Platform (kubernetes/docker/lambda/host…) and cloud (aws/gcp/…) can come from the repository or an unambiguous statement in the goal. For example, "my Kubernetes service on AWS" supplies platform=kubernetes and cloud=aws. Ask only when neither source determines the needed value, or when they conflict. Do not guess an unstated value or re-ask for one that is already clear.
  • Conduct, don't perform. You compose and dispatch skills; you never do a delegated skill's job, pre-empt its decisions, or turn them into a user choice. In particular: product selection is dd-product-recommender's job — never infer, guess, or shortlist products yourself, and never offer product scope as a choice; and authentication + site/region are dd-account-setup's job — invoke it and let it ask. The only choices you surface concern missing or conflicting structural context (platform/cloud), including the CHOICE POINTS that resolve.py emits.
  • Never fabricate a skill or a step; dispatch only what the resolver planned. If a recommended product has no node, or is not covered for the detected platform, say so plainly, point to docs.datadoghq.com, and record it as a gap. Two hard gates apply before any dispatch:
    • No plan, no dispatch. If you have not run resolve.py and captured its PLAN block (with the SESSION ID) for this run, dispatch nothing — the resolver plan is the only dispatch authority. Do not hand-build a plan.
    • Every id must be in the PLAN. Before you dispatch a skill id, confirm it appears in that PLAN block (e.g. grep the id in the run's trace file — see Step 5). An id not in the PLAN is fabricated: do not dispatch it; record it as a gap.
  • Account first. Nothing routes before dd-account-setup reports a valid key on the right region.
  • Invoke every planned skill — installed or not; never skip, never substitute. Dispatch each skill once, in the resolved order, from its recorded source in catalog.json. If the skill id is in the session registry, invoke it directly. If it is not, fetch it from source and run it inline (invoke, not install): python3 dd-orchestrator/scripts/fetch_skill.py <id> materializes the skill (its SKILL.md plus any references/ and scripts/) into a temp dir from the newest public source, and you then execute that SKILL.md. Sources always track the newest version (main / the live onboarding-API render) — nothing is pinned. You may NOT infer, summarize, or hand-author a skill's result in its place. The only permitted non-execution is a hard failure of the fetch or the skill itself, which stops that dependency chain and is reported (see Sequential dispatch below), or a node removed before the confirm gate because the developer's target is outside its CAVEATS: scope (see the PLAN block) — never a silent skip. If fetch_skill.py exits non-zero (for example, HTTP 404 or a blocked/non-allowlisted source), record that node under DEAD_ENDS with the actual fetch error and mark it not-run. Stop its dependency chain. Do not substitute a local, cached, vendored, or on-disk copy after a failed fetch, put it under DISPATCHED, or claim the product is set up.
  • Sequential dispatch — chains first, stop on failure. Run the plan strictly in resolve.py's order (it is deterministic and independent of the product request order, and keeps each dependency chain contiguous). Execute one skill at a time; do not start a skill until every hard prerequisite has succeeded, not merely been dispatched. Each skill runs at most once. If a skill fails, stop that chain: skip its transitive dependents and report them as not-run — an independent branch is unaffected.
  • Checklist discipline. Post a checklist up front and tick items as you go (per this repo's CLAUDE.md).
  • Output discipline — quiet by default. The only user-facing output is: the checklist (post it once, then tick items in place — do not reprint it), the PLAN block, any choice questions, and the SUMMARY block. Between the PLAN block and the SUMMARY block, write at most one short status line per plan node (e.g. -> dd-account-setup ... then done: dd-account-setup). Do not narrate your reasoning, read files aloud, restate the goal or the plan, or re-explain these rules. The audit detail belongs in the run's trace file (Step 5) — point to it; do not reprint it. Keep your reasoning internal; do not think out loud. Setup and telemetry plumbing — running preflight shell commands and emit.py record calls — is internal: run it silently and never announce it or report its outcome, unless the context explicitly asks for debug info. Fewer words, no lost meaning (this pairs with Simplified Technical English below).
  • Correctness first; styling is best-effort. The load-bearing path is: account first → compose with resolve.py → confirm → dispatch each planned skill once, in order, recording each boundary with emit.py record (it keeps the trace current, so it is not optional). The Simplified Technical English styling below is best-effort. Sending telemetry can fail without consequence: a failed send never changes record's exit code and the trace is still updated.
  • Prefer Simplified Technical English (ASD-STE100) (best-effort — see Correctness first above). All user-facing output — the checklist, the plan, choice questions, gaps, and the summary — uses Simplified Technical English (ASD-STE100). Apply its core rules:
    • Keep sentences short: at most 20 words for an instruction, 25 for a description. Write one instruction per sentence.
    • Use the active voice, the imperative for instructions, and simple verb tenses. Avoid -ing forms where a simpler verb works.
    • Keep the articles ("a", "the") and write complete sentences. Do not use a telegraphic or headline style.
    • Use one word for one meaning, and keep the same term for the same thing. Avoid slang, jargon, and undefined abbreviations.
    • Prefer short, common words. Use lists for steps, and keep paragraphs short. The result is short but complete: fewer words, no lost meaning.

The flow

developer goal
      │
      ├─►  dd-account-setup            precondition: valid key, right region
      ├─►  dd-product-recommender      goal + codebase → ranked PRODUCTS  (skipped if the intent names products)
      ├─►  detect context              platform + cloud from the repo or an unambiguous goal
      │
      ▼
  scripts/resolve.py --products "<recommended>" --platform <detected> --cloud <detected>
      │        (reads catalog.json: binds category requires to detected context,
      │         orders hard edges depth-first for dependency-chain locality
      │         (kind breaks ties), dedupes, appends verify)
      ▼
  PLAN (ordered skills) + DEAD-ENDS (recorded) + CHOICE POINTS (ask)
      │
      ▼
  confirm → dispatch each plan node by source.url → summarize (incl. the skipped gaps)

Steps

  1. Post the checklist.
  2. Ensure the account — invoke dd-account-setup (installed → invoke directly; not installed → fetch it from source and run it inline, per Ground rules); stop if it cannot produce a validated key. dd-account-setup owns the site/region and authentication prompts — do not pre-empt them; invoke it and let it ask. Leave the OAuth token it writes in place until the run ends: resolve.py (Step 5) uses it to read the org id for telemetry.
  3. Products — shortcut or recommend. First check whether the intent already names products: python3 dd-orchestrator/scripts/resolve.py --detect-products "<intent>".
    • If it prints one or more product tokens, the user named the products: use that list and skip the recommender. (e.g. "Instrument RUM and LLMO" → rum,llm-obs.)
    • If it prints nothing, the intent describes a goal: you MUST invoke dd-product-recommender and use its ranked product list (installed → invoke directly; not installed → fetch it from source and run it inline, per Ground rules). Never infer, guess, or shortlist products yourself from the stack/framework, and never offer product scope as a user choice — product selection is the recommender's job, and running it is mandatory here. (e.g. "Help me track user actions" → recommend.) The shortcut skips only the recommendation. Account setup (Step 2), context detection (Step 4), the confirm gate, and resolve.py still run — the shortcut is never a bypass of a safety gate. No legitimate products, no resolve.py. The product list passed to Step 5 MUST come from exactly one of: the --detect-products shortcut (→ --intent-mode explicit) or dd-product-recommender (→ --intent-mode recommended). You may not compose or preview a plan from products you authored yourself; resolve.py --trace refuses to run without a declared --intent-mode.
  4. Detect context — use repository evidence (k8s manifests, Dockerfile, serverless.yml, Terraform/SDK/provider signals) or an unambiguous platform/cloud stated in the goal. Pass known values to resolve.py as --platform / --cloud. Leave unknown values unset; ask when a needed value is missing or the repository and goal conflict. Do not replace an explicitly stated platform with none just because the repository lacks deployment files.
  5. Compose the plan and seed the trace. The trace is a run-scoped scratch file the orchestrator owns — not a bare output/ in the user's project. Ask the emitter for the path; it scopes the default under TMPDIR by the host agent's conversation/thread id (or by workspace as a fallback), so concurrent agents do not share a session pointer. DD_ORCH_OUTPUT_DIR overrides the directory. A host that exposes no session marker must set a unique DD_ORCH_OUTPUT_DIR for concurrent runs. Reset the scratch, then run: TRACE="$(python3 dd-orchestrator/scripts/emit.py trace-path)"; mkdir -p "$(dirname "$TRACE")" && rm -f "$TRACE" python3 dd-orchestrator/scripts/resolve.py --trace --products "<products>" --platform <platform> --cloud <cloud> --intent-mode <explicit|recommended> | tee "$TRACE" Pass --intent-mode explicit when Step 3's shortcut named the products, or --intent-mode recommended when dd-product-recommender produced them. It is required — resolve.py --trace refuses to compose a plan without it (the products must trace to the shortcut or the recommender, never to your own inference) — but it never changes the plan itself. The --trace flag prints one stable, machine-readable block — SESSION_ID, STOP_REASON, PLAN, DEAD_ENDS, CHOICE_POINTS, SUGGESTED, CONFIRMED, DISPATCHED — and tee saves it verbatim to $TRACE. That deterministic block IS your dispatch trace; never re-narrate the plan by hand. Read the same output for the ordered plan, the dead-ends, and any choice points. Capture the SESSION_ID: value from that block in a run-local variable: SID=<SESSION_ID>. Pass --session-id "$SID" to every later resolve.py re-run. Capture a fresh value from the first trace of each new run; do not export DD_ORCH_SESSION_ID. If an outer wrapper supplies that environment variable, the resolver honors it; leave its value unchanged. resolve.py also saves the id to a state file, so emit.py record needs no id, and a re-run whose $SID is empty in a new shell stays in the same run. A new id during a replan would split one run's telemetry and restart its event_seq. The saved session is scoped to the same directory as $TRACE, and every record call verifies that the trace's embedded session id matches before changing state. In debug mode (only when the context explicitly asks for it), add --debug to also render the ASCII DAG (indent = dependency depth, <- = direct prerequisites). resolve.py also emits the plan telemetry here (see Run records below). Immediately after the first trace, record the completed dd-account-setup preflight exactly once with record started dd-account-setup then record finished dd-account-setup --result success. These post-hoc boundary records deliberately omit account duration; they do not pretend to time work completed before the session existed. A choice-driven replan restores successful dispatches from state automatically; do not record the account again.
  6. Resolve choices — for each choice point (e.g. "pick a platform: kubernetes, linux"), ask the developer and re-run the Step 5 command with the confirmed value, the same --intent-mode, and --session-id "$SID" so the updated plan and subsequent telemetry remain part of the same run. Write the output to a new file and replace $TRACE only when the command succeeds, so a failed re-run cannot empty the trace: python3 dd-orchestrator/scripts/resolve.py --trace ... --session-id "$SID" > "$TRACE.new" && mv "$TRACE.new" "$TRACE" && cat "$TRACE"
  7. Confirm — render the PLAN block, then dispatch. Before any dispatch, render the PLAN block (see Output templates below) verbatim: fill the slots, add no extra prose, keep the exact section order and headers. The Plan table's Source column is the skill's source URL, as resolve.py prints it (the newest public source; self for this repo). On approval, run each plan node in dispatch order, one at a time — invoking each skill (installed → directly; not installed → fetched from source and run inline, per Ground rules) — and ticking the checklist. Whenever the plan begins with dd-account-setup (every plan that needs an account — i.e. any non-empty onboarding plan), Step 2 already ran it: do not invoke it a second time; Step 5 already recorded it, so tick that node, retain its first DISPATCHED entry, and continue with the next node. Do not emit another account lifecycle event. Do not start a node until its prerequisites succeeded; if one fails, skip its dependents. Keep the trace file ($TRACE, Step 5) current with emit.py record (Run records below): record confirmed on approval or record declined if the user declines; record started <id> before each node; then record finished <id> --result success|failed, or record skipped <id> for a node that does not run. record finished --result success adds the id to DISPATCHED once, in execution order, so DISPATCHED lists only skills that ran successfully, directly if installed or fetched-and-run otherwise. A failed fetch or execution belongs under DEAD_ENDS with its actual error; distinguish "not run" from "ran and failed", and stop that dependency chain. DISPATCHED records successful executions, starting with the account preflight. If choices remain, the plan is empty, or the user declines, run no further skills; keep the successful preflight entry without inventing additional dispatches. After record declined, go to Step 8: record run-finished closes every unrun node as declined. A node removed under the scope rule (see the PLAN block) is not run: record it under DEAD_ENDS as - <skill_id> — out of scope: <the caveat it falls outside>, keep it out of DISPATCHED, and close it with record skipped <skill_id> --result out_of_scope at the confirm gate, before you start any later node (a later node cannot start while an earlier node is open).
  8. Summarize — render the SUMMARY block. As the terminal output, render the SUMMARY block (see Output templates below) verbatim, in the exact section order. The skipped / gaps list is a real output — the coverage-gap / demand signal for what to automate next; state it, do not hide it. Run emit.py record run-finished once (Run records) — except while a choice question from Step 6 is waiting for the developer's answer: the run is paused, not finished, so do not run it then; finish the run after the replan.

Output templates

Render two fixed blocks so every run reads the same: the PLAN block at Step 7 (the confirm gate) and the SUMMARY block at Step 8 (terminal output). Fill the slots and add no extra prose.

Rendering rules (both blocks). Use the given section order and the exact headers. Prefer tables to prose; one line per row; do not editorialize or restate the goal. Every dispatched step shows its source (the skill's public source URL — newest; self for this repo). Action items are imperative and carry the exact command or file path. Always end with the telemetry session_id so the output and the events join. Omit a section only by its stated omission rule. Marker legend: ✓ done/verified · ◑ partial or wired-not-verified · ⚠ needs action / mutating · ✗ failed/blocked · ⊘ skipped/not-covered.

PLAN block — Step 7 (before any dispatch)

# Datadog Onboarding — Plan · session {{session_id}}

## Detected context
- Platform: {{platform}}  ·  Cloud: {{cloud|none}}  ·  Stack: {{stack_summary}}
- Existing Datadog: {{existing|none}}

## Recommended products
{{i}}. {{product}} · {{priority}} · {{one-line why, names a file/lib}}

## Plan — {{k}} step(s), in dispatch order
| # | Skill | Kind | Product | Source |
|---|-------|------|---------|--------|
| {{n}} | {{skill_id}} | {{kind}} | {{product}} | {{source url (newest) or self}} |
Dependencies: {{root}} → {{chain / branches, one line}}

## Not automated ({{dead_end_count}})
- {{product}} — {{why}} → {{docs URL}}
> Show "None — every recommended product is covered." when dead_end_count = 0.

## Decisions needed ({{choice_count}})
- {{choice}}: {{optionA}} / {{optionB}} / {{optionC}}
> Show "None." when choice_count = 0.

## Scope limits ({{caveat_count}})
- {{skill_id}}: {{caveat}}
> Omit this section when the trace has no `CAVEATS:` block.

## Before you approve — effects
- ⚠ {{step}} {{mutating / outward-facing effect}}
- {{step}} {{non-mutating effect}}

Approve?  [Proceed — all {{k}}]  ·  [Cancel]

When the trace has a CAVEATS: block, check each caveat against what the developer asked for and the detected context before the confirm gate. If their target is outside a skill's caveat (for example a MySQL or Aurora database for an RDS PostgreSQL skill, or a serverless service for the profiler), do not dispatch that skill or any prerequisite that only it needs; list the product under "Not automated" instead, drop those nodes from the Plan table, and record them in $TRACE as Step 7 describes.

SUMMARY block — Step 8 (terminal output)

# Datadog Onboarding — Summary · session {{session_id}} · {{result}}

## Checklist
- [{{x|.}}] #{{n}} {{skill_id}} — {{one-line outcome}}

## Products set up
| Product | Delivered by | Status | Evidence |
|---------|--------------|--------|----------|
| {{product}} | {{mechanism, source}} | {{✓|◑|✗}} | {{proof or "pending {{blocker}}"}} |

## Changed
- Cluster: {{namespaces/resources}}
- App: {{files/manifests}}
- Creds: {{where, gitignored?}}

## Action items — do next
1. [ ] {{imperative}} — `{{exact command / path}}`
> Show "None — setup is complete." when there are no follow-ups.

## Issues & deviations
| What | Cause | Resolution / impact |
|------|-------|---------------------|
| {{issue}} | {{cause}} | {{how resolved / residual impact}} |
> Show "None." when the run was clean.

## Gaps / demand signals
- {{skipped product or catalog/orchestrator gap}}

## Verify in Datadog
- {{product}}: {{deep link}}

## Telemetry
session {{session_id}} · {{event_count}} events · result {{result}} ({{s}}✓ / {{f}}✗ / {{k}}⊘)

Run verdict — the overall {{result}}: use success when every planned skill succeeded, even if some recommended products are dead-ends / "Not automated" (no skill yet) — those are coverage gaps, not partial failures. Use partial_success when a planned fetch or execution failed, or a dependent step was skipped; do not infer success from DISPATCHED alone. Use blocked when nothing beyond the account preflight can run because every product dead-ended; cancelled when the user declined the plan. Record failures and skips with emit.py record even though those skills are absent from DISPATCHED; emit.py reconciles the verdict from those counts.

Run records (emit.py record) and telemetry

Every dispatch boundary is one emit.py record call. It updates the trace file ($TRACE) and sends the matching telemetry event in the same step, so you never edit CONFIRMED or DISPATCHED by hand and never call the emitter separately. record takes the plan, the session id, and the step fields from the state that resolve.py saved, so a fresh shell needs no $SID. It also computes duration_ms and the final step counts itself.

python3 dd-orchestrator/scripts/emit.py record confirmed          # user approved the plan -> CONFIRMED: yes
python3 dd-orchestrator/scripts/emit.py record declined           # user declined -> CONFIRMED: no
python3 dd-orchestrator/scripts/emit.py record started  <skill_id> --source-mode <installed|fetched>
python3 dd-orchestrator/scripts/emit.py record finished <skill_id> --result <success|failed>   # success -> DISPATCHED
python3 dd-orchestrator/scripts/emit.py record skipped  <skill_id> [--result skipped_dependency|declined|source_unreachable|out_of_scope]
#   default: skipped_dependency (a prerequisite failed) or declined; use source_unreachable when the fetch
#   failed and out_of_scope for a node removed under the CAVEATS scope rule
python3 dd-orchestrator/scripts/emit.py record run-finished       # once, at the end

record run-finished derives result from the recorded steps: cancelled after record declined (it also closes the unrun nodes as declined), and blocked when nothing beyond account preflight ran because every product dead-ended. Pass an explicit terminal result only when that derivation would be wrong. A run cannot continue after record run-finished, so never run it while a choice question is waiting for the developer.

Valid commands exit 0 even when sending fails; malformed commands, unknown skill ids, trace/session mismatches, and duplicate or out-of-order transitions exit 2 without changing the trace, and print the reason on stderr; correct the call from that reason instead of retrying it unchanged. Sending is best-effort (bounded timeout, local debug log, never throws), goes to the logs-intake route only, and is turned off with DD_ORCH_TELEMETRY_DISABLED=1; valid trace updates still happen. Never build your own HTTP request or curl. Automated runs set DD_ORCH_RUN_CONTEXT (eval, ci or dev) so they are kept apart from real usage; without it, a run is ci when CI is set and user otherwise.

resolve.py emits the plan core on each resolve: one skill_run:started per session, then skill_run:plan_resolved and one skill_step:planned per plan node and per dead-end. Each resolve increments plan_revision; all planned events for that revision form one complete snapshot (each skill_step also carries depends_on, the CSV of prerequisite plan positions, so the DAG edges are reconstructable). It also saves the run envelope (agent, run context, platform, cloud, entry, intent mode, org id), and emit.py re-attaches it and an emitted_at (ms) timestamp to every event. The agent is detected from its environment. Org lookup uses the validated API/app-key pair from the environment or parsed .env.local / .env and the run-owned copy of the OAuth token. If both identify different organizations, it omits org_id instead of guessing. The emitter refuses unknown sites before sending credentials. record run-finished releases this run's token claim and removes the shared handoff token only after the final concurrent claim ends.

Field values are bounded enums / ids / counts only — never send goal text, paths, keys, URLs, or model output (the emitter also strips anything not on its allow-list).

Reliability and reconciliation. Every event carries a per-session event_seq (a monotonic ordinal): a gap in event_seq means an event was dropped, not that the step never ran. resolve.py emits its plan-shape core (skill_run:started, skill_run:plan_resolved, one skill_step:planned per node) as critical — one transport blip cannot drop the whole core. emit.py also appends every attempted event to a durable local NDJSON log for offline audit. When you analyze a run, treat step_success_count / step_failed_count / step_skipped_count on skill_run:finished as the source of truth for how many steps ran; reconcile the per-step events against it. Do not assume a missing per-step event means the step did not run. emit.py also reconciles the terminal skill_run:finished result against those counts (and the dead-end count): if the reported verdict contradicts them, it keeps the reported value as result_reported and sets result to the count-consistent verdict (result_reconciled: true). So report the honest per-step results and let the guard settle the run verdict.

Worked example

Goal "monitor my Node service on Kubernetes" → recommender [APM, Infrastructure Monitoring], detected platform=kubernetes:

1. dd-account-setup            (foundation)          [self]
2. apm-agent-install-kubernetes (platform-install)   [agent-skills]
3. apm-enable-kubernetes       (product-enable/apm)  [agent-skills]
4. apm-verify-ssi-kubernetes   (verify)              [agent-skills]

One product plus a detected platform became a four-skill plan drawn from two repos, correctly ordered, with Infrastructure Monitoring delivered by the same SSI Agent install. The intent never entered the resolver — only the products and the detected platform did.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.