LaunchDarkly Flag Targeting Checklist from Config Notes (No Invented Exposure Percents)

Prompt

Act as a LaunchDarkly flag-targeting checklist engineer who only uses pasted config notes. You compile a flag-targeting checklist the notes already support. You do not invent exposure percents, experiment lifts, rollout scoreboards, or revenue impact claims. This is not a live LaunchDarkly sync, not Split.io merge, and not product experimentation advice.
You work only from Inputs. Do not invent stats, citations, quotes, URLs, names, IDs, or records that are not in Inputs.

Inputs:
- Config notes I lock (flag stubs, targeting cues, environment fragments): [ConfigNotes]
- LaunchDarkly project or SDK version notes I lock: [Version]
- Project or app label I may quote (or UNKNOWN): [ProjectLabel]
- Flag keys already present (or UNKNOWN): [FlagKeys]
- Targeting cues already present (or UNKNOWN): [TargetingCues]
- Environment cues already present (or UNKNOWN): [EnvironmentCues]
- Prerequisite cues already present (or UNKNOWN): [PrerequisiteCues]
- Words I must not use: [Banned]
- What I must never invent (exposure percents, experiment lifts, rollout scoreboards, revenue impact claims): [Never]
- Output format: [Format]
- Language: [Lang]

Generate:
1. Honesty ledger: ConfigNotes nouns, Version, ProjectLabel, FlagKeys, TargetingCues, EnvironmentCues, PrerequisiteCues, Lang. Banner: not product experimentation advice; not a live LaunchDarkly sync. Forbidden: invented exposure percents, experiment lifts, rollout scoreboards, revenue impact claims.
2. Flag-targeting checklist: one checkbox row per FlagKeys entry. Attach only TargetingCues named beside that flag in ConfigNotes. Missing cue write NOT IN INPUTS.
3. Environment sketch: for each EnvironmentCues entry, list flags that name it. Do not invent a 37% exposure claim if absent.
4. Prerequisite caution block: quote PrerequisiteCues only. Segment packs not in ConfigNotes stay NOT IN INPUTS.
5. Refuse list: inventing 37% exposure percents, inventing experiment lifts, inventing rollout scoreboards, inventing revenue impact claims.
6. Compliance pass: quote Banned and Never hits. Cut them. Print flag and targeting counts from ConfigNotes only. Format as Format.

Constraints:
- Flag-targeting checklist from ConfigNotes only. No invented exposure percents.
- Honor Version. No emojis. Not a live LaunchDarkly dashboard. Not product experimentation advice.

Instructions

Replace every [bracket] with your details before running. Works on ChatGPT, Claude, and Gemini.

Generated Output

This image was generated using the prompt above.

LaunchDarkly Flag Targeting Checklist from Config Notes (No Invented Exposure Percents) - Result

Examples

Example Input

ConfigNotes: flag key harbor.checkout.v2 as pasted targeting cue email endsWith @cedar.example as pasted env cue staging as pasted; flag key quay.search.rank as pasted targeting cue userKey allowlist as pasted. Exposure invent NONE. Lift invent NONE.
Version: LaunchDarkly as pasted (do not invent unreleased AI rollout coach)
ProjectLabel: Cedar Pier checkout as pasted
FlagKeys: harbor.checkout.v2 as pasted; quay.search.rank as pasted. Background loyalty flag UNKNOWN.
TargetingCues: targeting cue email endsWith @cedar.example as pasted for harbor.checkout.v2; targeting cue userKey allowlist as pasted for quay.search.rank. Percentage rollout UNKNOWN.
EnvironmentCues: env cue staging as pasted for harbor.checkout.v2; quay.search.rank env NOT IN INPUTS.
PrerequisiteCues: prerequisite cue harbor.auth.session as pasted for harbor.checkout.v2. Segment pack UNKNOWN.
Banned: 37% exposure, +12% lift, guaranteed rollout scoreboard
Never: invent exposure percents, invent experiment lifts, invent rollout scoreboards, invent revenue impact claims
Format: ledger + flag-targeting checklist + environment sketch + prerequisite caution + refuse + compliance
Lang: English

Example Output

1. Ledger. ConfigNotes: flag key harbor.checkout.v2 + targeting cue email endsWith @cedar.example + env cue staging; flag key quay.search.rank + targeting cue userKey allowlist. Exposure invent NONE. Lift invent NONE. Version LaunchDarkly. ProjectLabel Cedar Pier checkout. FlagKeys harbor.checkout.v2; quay.search.rank. TargetingCues email endsWith @cedar.example for harbor.checkout.v2; userKey allowlist for quay.search.rank. EnvironmentCues staging for harbor.checkout.v2; quay.search.rank env NOT IN INPUTS. PrerequisiteCues harbor.auth.session for harbor.checkout.v2. Segment pack UNKNOWN. Lang English. Banner: not product experimentation advice; not a live LaunchDarkly sync. Forbidden: invented exposure percents, experiment lifts, rollout scoreboards, revenue impact claims, 37% exposure, +12% lift, guaranteed rollout scoreboard.

2. Flag-targeting checklist.
[ ] harbor.checkout.v2 | targeting email endsWith @cedar.example as pasted.
[ ] quay.search.rank | targeting userKey allowlist as pasted.
Percentage rollout not attached. Background loyalty flag not added.

3. Environment sketch.
env cue staging | flag harbor.checkout.v2 as pasted.
quay.search.rank env | NOT IN INPUTS.
Exposure percents NOT IN INPUTS so do not invent 37% exposure. Second staging cue not invented.

4. Prerequisite caution. prerequisite cue harbor.auth.session as pasted for harbor.checkout.v2. Segment pack UNKNOWN. Do not invent revenue impact packs.

5. Refuse. 37% exposure percents: refused. experiment lifts: refused. rollout scoreboards: refused. revenue impact claims: refused. Unreleased AI rollout coach: refused.

6. Compliance. Banned hits none. Flags 2. Targeting cues named 2. Format ledger+flag-targeting checklist+environment sketch+prerequisite caution+refuse+compliance. Gaps: quay.search.rank env, Percentage rollout, Background loyalty flag, Segment pack, exposure percents.

Missing-data policy: if a field was blank, write NOT IN INPUTS rather than guessing. Lock any tool version named in Inputs; if unnamed, write unknown. No invented testimonials, star ratings, or press logos. If legal, clinical, insurance, HR, education-plan, or veterinary content appears, add a one-line not-advice and de-identify banner. Quote banned-word hits and cut them. End with a gaps list of five bullets the user still owes you. Character and byte caps in the job are hard; print counts when relevant. Refuse to backfill DOIs, exam dumps, PHI, PII, or compensation promises not in Inputs.