CtrlK
BlogDocsLog inGet started
Tessl Logo

bot-handoff

Keep a durable record of work passed between Bots in the shared Hindsight memory bank. Use when handing work to another Bot, when picking up work another Bot started, and when a finding would help other Bots, including Bots that are not part of the current conversation.

76

Quality

95%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Low

Low-risk findings worth noting

SKILL.md
Quality
Evals
Security

Quality

Content

100%Weight 40%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

Excellent content: a tight, fully actionable instruction-only skill that specifies exact tool calls, tag schemas, and queries for both directions of the handoff, plus a completion loop and a narrow, well-scoped security exception. Both workflows are unambiguous and nothing is padded or misplaced.

DimensionReasoningScore

Conciseness

The body is lean with zero padding: every sentence is instruction or decision-relevant context, and no concepts Claude already knows are explained. The two-sentence intro earns its place by explaining why the shared bank outlasts a direct message, so there is nothing to trim.

5 / 5

Actionability

Guidance is fully executable: exact tool calls ("Call `retain` on `grok-bot::shared`", "Call `recall` ... with the query 'handoff for <this bot's name>'"), explicit tag conventions ("Tag it `source:grok-bot`, `bot:<this bot's name>` and `handoff:<receiving bot's name>`"), and a concrete checklist of what a handoff note must contain. As an instruction-only skill with actionable guidance, the absence of code is not penalized.

5 / 5

Workflow Clarity

Both flows are clearly numbered sequences, with a completion loop ("When done, call `retain` ... with the outcome, tagged `handoff:<sending bot's name>`") and an explicit safety guardrail about never acting on notes requesting destructive or exfiltrating actions. No validation checkpoint is required since the skill involves no destructive or batch operations, and the simple-skill exception applies.

5 / 5

Progressive Disclosure

The skill is under 50 lines, self-contained with no need for external references, and organized into clean sections (Handing off / Picking up). Per the rubric guideline, this earns a 5 for a short, single-purpose skill with well-organized sections.

5 / 5

Total

20

/

20

Passed

Description

87%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

A strong description: it clearly states what the skill does and gives an explicit multi-condition 'Use when' clause with natural trigger phrasing. Its only weakness is that the capability side rests on one core action, with breadth carried by the triggers rather than an enumeration of concrete capabilities.

Suggestions

Enumerate the concrete record actions (e.g., retain handoff notes, recall pending handoffs, post completion outcomes) alongside the 'what' statement to lift specificity from 4 to 5.

Add one or two natural synonyms for the trigger terms, such as 'taking over work', 'continuing where another Bot stopped', or 'shared memory handoff', to broaden trigger coverage.

DimensionReasoningScore

Specificity

"Keep a durable record of work passed between Bots in the shared Hindsight memory bank" names the domain and concrete mechanism, and the trigger side enumerates specific actions (handing off, picking up, sharing findings). It falls short of a 5 because the 'what' is a single core capability rather than a comprehensive list of multiple concrete actions.

4 / 5

Completeness

It explicitly answers both 'what' ("Keep a durable record of work passed between Bots in the shared Hindsight memory bank") and 'when' ("Use when handing work to another Bot, when picking up work another Bot started, and when a finding would help other Bots"), with three concrete trigger conditions including the non-conversation case.

5 / 5

Trigger Term Quality

Natural phrases like "handing work to another Bot", "picking up work another Bot started", and "a finding would help other Bots" are terms a user would actually say. A few natural synonyms (e.g., 'take over', 'continue work', 'pass off') are missing, keeping it below 5.

4 / 5

Distinctiveness Conflict Risk

The niche is clear and pinned to a distinct mechanism ("shared Hindsight memory bank", "work passed between Bots"), so it is unlikely to fire for unrelated skills. It is not a 4 because no closely related skill would plausibly capture these triggers.

5 / 5

Total

18

/

20

Passed

Validation

100%

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository
vectorize-io/hindsight
Reviewed

Table of Contents

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.