negative-targeting
Skill8 starsChanged 53 days agoMITView on GitHub
Generate action-safe Amazon Ads negative keyword or negative product-targeting candidates with protected-term, origin/context and collateral-damage checks. Use when the primary task is choosing negative type/scope; route Search Term row interpretation to search-term-analysis.
In this repository16 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- amazon-ads-auditSkill
- amazon-ads-optimizerSkill
- anomaly-detectionSkill
- bid-optimizationSkill
- budget-optimizationSkill
- campaign-health-monitorSkill
- experiment-plannerSkill
- growth-opportunity-finderSkill
- keyword-optimizationSkill
- performance-drop-diagnosisSkill
- placement-optimizationSkill
- post-change-reviewSkill
- profitability-analysisSkill
- search-term-analysisSkill
The file
We split this skill into the rules it gives an agent
3 rules. About 0 are common, 1 is distinctive, and 2 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈1
Distinctive · estimate
This author's own choices. Worth a look.
2
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 18
When a conclusion depends on live MCP/API/connector data, load ../../references/connector-capability.md before interpreting missing or …
- Very specific: line 36
Search Term row 缺少 origin/context/control mapping 时,不因为用户说“否词”就跳过解释层;先补证据或保持 Manual Review。
- Distinctive: line 41
Search Term 或 product targeting 性能;
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 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.
Files
Negative Targeting
Connector capability gate
When a conclusion depends on live MCP/API/connector data, load ../../references/connector-capability.md before interpreting missing or empty fields. If a machine-readable capability snapshot is available, identify the exact required capability IDs and run ../../scripts/evaluate_connector_capability_gate.py before metric interpretation.
Pass— continue with the Skill's normal evidence, sufficiency, lineage and safety checks.Degraded— do not make a high-confidence dependent recommendation; keep the result toDirectional,Hold,Alternate Source,Missing Data, orManual Review.Blocked— do not treat the dependent observation as action-safe until the capability is resolved or an allowed alternate source is verified.missing_evidence_policy = never_zero—Partial,Unsupported,Unknown, or absent connector capability is never a numeric zero, unchanged state, or proof that the platform lacks the feature.
This gate is read-only and does not authorize live Amazon Ads mutation.
默认模式
Suggest。否词属于可能直接切断流量的高影响动作,默认要求人工复核。
Routing boundary
- 本 Skill 负责 negative control decision:选择 Negative Exact / Negative Phrase / Negative Product Targeting 的类型与 scope,并检查 protected terms、origin/context、历史赢家、归因延迟和 collateral-damage risk。
- 如果主要任务是解释 Search Term report row、判断
term_origin、report coverage、query/traffic quality、winner/waste evidence,先路由到../search-term-analysis/SKILL.md;本 Skill 消费该证据,不复制 Search Term 解释层。 - 如果主要任务是 configured keyword/target 的结构、match type、state 或 lifecycle,而不是否定控制,路由到
../keyword-optimization/SKILL.md。 - Search Term row 缺少 origin/context/control mapping 时,不因为用户说“否词”就跳过解释层;先补证据或保持
Manual Review。
输入
需要:
- Search Term 或 product targeting 性能;
- 来源 campaign/ad group/target;
- clicks、spend、orders、sales;
- 日期范围;
- 业务目标。
强烈推荐:
term_origin:literal_query/inferred_non_search/unknown;- placement / traffic context / matched targeting evidence;
- 品牌词保护表;
- 核心类目词/战略词表;
- 历史赢家;
- 已存在 Exact 收割结构;
- 归因延迟和促销上下文。
Amazon 当前 Sponsored Products Search Term report 可能包含 shopper search query,也可能包含 non-search context 中由系统推断出的 best match。因此 Search Term row 的显示字符串不自动等于 literal shopper query。
候选类型
Negative Exact
适合只屏蔽一个明确查询,误伤范围较小。对于单个低效/不相关 query,优先考虑 Exact 而不是 Phrase,但前提是 term_origin / context 足以证明显示字符串确实对应可由 Negative keyword 控制的 literal shopper query。
Negative Phrase
影响范围更广。只有当一个短语意图整体不相关、origin/context 明确且已评估长尾误伤时才建议。
Negative Product Targeting
适用于明确低价值 ASIN/Category traffic,需确认广告类型和 API/控制台是否支持对应否定能力。
判断顺序
- 该 row 是 literal shopper query、inferred non-search match,还是
unknownorigin? - 实际可控对象是 keyword、product target、placement/context,还是尚不明确?
- 查询/目标是否相关?
- 是否已有足够样本?
- 是否存在历史转化?
- 是否为品牌、防御、战略或高价值 NTB 流量?
- 是否可能存在归因延迟?
- 是否处于大促/价格变化窗口?
- 是否应该先 Exact 收割赢家再否定源流量?
- Negative Phrase 是否会覆盖其他有价值 query?
任何关键问题无法确认时,标记 requires_human_review: true。
如果 term_origin = inferred_non_search 或 unknown,不要仅依据显示字符串输出 action-safe Negative keyword。先寻找 matched target / placement / traffic-context 证据;无法建立控制映射时保持 Manual Review。
0 订单处理
不使用单一固定点击阈值。结合:
- 历史 CVR;
- 当前 CPC;
- 商品价格;
- target CPA/ACOS;
- 点击样本;
- 查询相关性;
term_origin与 traffic context;- 时间跨度。
如果 literal query 明显语义不相关,即使样本不大也可以列为“高相关性风险”候选,但仍默认人工复核。对于 inferred non-search / unknown-origin rows,字面语义不相关本身不足以证明 Negative keyword 是正确控制。
收割后否定
当 Search Term 是赢家但来自 Broad/Phrase:
- 先确认
term_origin足以支持 literal-query interpretation; - 再建议建立 Exact;
- 确认 Exact 处于 enabled 且可承接流量;
- 再根据结构目标决定是否对源入口 Negative Exact;
- 不先否定后创建,避免流量断层。
输出
每个候选输出:
- search term / target;
term_origin/ origin confidence;- proposed negative type;
- source entity;
- spend/clicks/orders/sales;
- reason;
- evidence;
- protected-term checks;
- collateral-damage risk;
- confidence;
- mode: Suggest;
- requires_human_review;
- rollback/recovery note。
禁止
- 不把 inferred non-search / unknown-origin Search Term row 仅按显示字符串自动转成 Negative keyword;
- 不批量否定品牌词而不提示;
- 不将所有 0 订单词直接否定;
- 不用 Negative Phrase 代替精细诊断;
- 不声称否词一定会降低 ACOS;
- 不在不知道现有 Exact/流量结构时自动做“收割 + 否词”执行计划。
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
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.

